{"schemaVersion":1,"qualityRubric":{"id":"svg-visual-v2","title":"StealthMark SVG score","description":"Benchmark performance across eight categories.","categories":[{"id":"fidelity","label":"Task & likeness","description":"Subjects, actions and likeness."},{"id":"structure","label":"Anatomy & construction","description":"Coherent bodies, joints and machinery."},{"id":"space","label":"Space & placement","description":"Perspective, scale and contact."},{"id":"light","label":"Light & reflections","description":"Lighting, shadows and reflections."},{"id":"detail","label":"Purposeful detail","description":"Clear, purposeful small features."},{"id":"artistry","label":"Artistry & character","description":"Composition, expression and drawing skill, including complexity and its execution."},{"id":"effects","label":"Materials & effects","description":"Surfaces, texture and movement."},{"id":"integrity","label":"Visual integrity","description":"Clean shapes, layers and edges."}]},"groups":[{"id":"selfie","name":"Selfie","description":"Visual construction\nand creative interpretation","taskComplexity":{"cognitive":2,"visual":4,"scale":10},"capabilityProfile":{"primary":["visual"],"secondary":["general"],"focus":"Character construction, expression and visual interpretation."}},{"id":"mona-lisa","name":"Mona Lisa","description":"Precise visual recall\nand portrait reconstruction","taskComplexity":{"cognitive":2,"visual":8,"scale":10},"capabilityProfile":{"primary":["visual"],"secondary":["general"],"focus":"Precise visual recall, likeness, anatomy and painterly reconstruction."}},{"id":"beach","name":"Beach","description":"Creativity, perspective\nand visual beauty","taskComplexity":{"cognitive":1,"visual":5,"scale":10},"capabilityProfile":{"primary":["visual"],"secondary":["general"],"focus":"Scene composition, beauty, perspective, placement and scale."}},{"id":"newtons-mirror","name":"Newton's Mirror","description":"Physical, spatial\nand recursive reasoning","taskComplexity":{"cognitive":7,"visual":7,"scale":10},"capabilityProfile":{"primary":["general"],"secondary":["visual"],"focus":"Chained physics, perspective and recursive reflection geometry."}},{"id":"ground-truth","name":"Ground Truth","description":"Physical reasoning, light\nand visual reconstruction","taskComplexity":{"cognitive":8,"visual":5,"scale":10},"capabilityProfile":{"primary":["general","visual"],"secondary":[],"focus":"Physical relationships, source-dependent colour and independent visual execution."}},{"id":"raccoon","name":"Raccoon","description":"Scene composition\nand interacting subjects","taskComplexity":{"cognitive":2,"visual":4,"scale":10},"capabilityProfile":{"primary":["visual"],"secondary":["general"],"focus":"Scene composition, anatomy and the spatial relationships between subjects."}},{"id":"pelican","name":"Pelican","description":"Subject construction and spatial relationships","taskComplexity":{"cognitive":0,"visual":3,"scale":10}},{"id":"foresight-self-improvement","name":"Foresight","description":"Anticipating threats\nand planning ahead","kind":"game-benchmark","family":"foresight","preview":"/shared/v1/game-preview-foresight-self-improvement.png","inContextLearning":true,"taskComplexity":{"cognitive":7,"scale":10,"learning":2},"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"previewScore":{"score":1,"turns":0,"limit":28}},{"id":"latch-five-levels","name":"Latch","description":"Cooperative planning\nand learning from experience","kind":"game-benchmark","family":"latch","preview":"/shared/v1/game-preview-latch-five-levels.png","inContextLearning":true,"taskComplexity":{"cognitive":7,"scale":10,"learning":4},"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"previewScore":{"score":2.4,"turns":0,"limit":12}}],"models":[{"id":"pixel-canary","identityStatus":"revealed","key":"pixel-canary","name":"Pixel Canary","revealedName":"Callstack Apex","aliases":["pixel-canary","stealth/pixel-canary","Apex","Callstack Apex","callstack/apex","apex-20261001"],"group":"Revealed models","url":"https://stealthmodels.com/pixel-canary","route":"callstack/apex","providerUrl":"https://vercel.com/ai-gateway/models/apex","license":{"label":"Proprietary hosted service","url":"https://apex.callstack.com/models","access":"proprietary"},"artwork":"pixel-canary","favicon":"/_shared/favicons/pixel-canary.png","openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":49.6,"rawScore":49.6,"count":8,"countedResults":7,"prompts":7,"bySubject":{"selfie":{"score":74.1,"rawScore":74.1,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":45.9,"rawScore":45.9,"count":1,"countedResults":1,"prompts":1},"beach":{"score":61.4,"rawScore":61.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":32.7,"rawScore":32.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":44.1,"rawScore":44.1,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":77.6,"rawScore":77.6,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":76.3,"rawScore":76.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"latch-five-levels":{"score":100,"rawScore":100,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":40.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":28.4,"upper":57.9},"availableFamilies":7,"expectedFamilies":8,"missingFamilies":["foresight-self-improvement"],"partial":true},"visual":{"score":59.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":57.6,"upper":61.8},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":42},"multiStep":{"score":50}}},{"id":"space-bunny-alpha","identityStatus":"revealed","key":"space-bunny-alpha","name":"Space Bunny Alpha","revealedName":"MiniMax M3.1 Flash (early version)","aliases":["space-bunny-alpha","Space Bunny Free","Space Bunny","stealth/space-bunny-alpha"],"group":"Revealed models","url":"https://spacebunnyalpha.com/","route":"stealth/space-bunny-alpha","providerUrl":"https://openrouter.ai/stealth/space-bunny-alpha","license":{"label":"Proprietary hosted service","url":"https://www.minimax.io/blog/minimax-m31f","access":"proprietary"},"artwork":"space-bunny-alpha","favicon":"/_shared/favicons/space-bunny-alpha.png","openModel":false,"localCapable":false,"allowPreviousGenerations":true,"defaultResults":["space-bunny-alpha-svg-raccoon-6ee169059ae55c1a"],"defaultEdition":"selected-defaults","scoreSummaries":{"from-20260929":{"score":45.4,"rawScore":45.4,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":78.9,"rawScore":78.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":49.9,"rawScore":49.9,"count":1,"countedResults":1,"prompts":1},"beach":{"score":67.8,"rawScore":67.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":24.9,"rawScore":24.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":42.7,"rawScore":42.7,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":73.4,"rawScore":73.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":68,"rawScore":68,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":35.8,"rawScore":35.8,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":34.4,"rawScore":34.4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":35.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":35.2,"upper":36.2},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":64.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":62.1,"upper":66.4},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}},"before-20260929":{"score":54.9,"rawScore":54.9,"count":7,"countedResults":6,"prompts":6,"bySubject":{"selfie":{"score":80.1,"rawScore":80.1,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":48.1,"rawScore":48.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":75.6,"rawScore":75.6,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"ground-truth":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"raccoon":{"score":78.4,"rawScore":78.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":79.8,"rawScore":79.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":35.8,"rawScore":35.8,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":34.4,"rawScore":34.4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":43.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":20.5,"upper":73.1},"availableFamilies":6,"expectedFamilies":8,"missingFamilies":["newtons-mirror","ground-truth"],"partial":true},"visual":{"score":67.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":48.1,"upper":78},"availableFamilies":4,"expectedFamilies":6,"missingFamilies":["newtons-mirror","ground-truth"],"partial":true}}},"selected-defaults":{"score":45.8,"rawScore":45.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":78.9,"rawScore":78.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":49.9,"rawScore":49.9,"count":1,"countedResults":1,"prompts":1},"beach":{"score":67.8,"rawScore":67.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":24.9,"rawScore":24.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":42.7,"rawScore":42.7,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":78.4,"rawScore":78.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":68,"rawScore":68,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":35.8,"rawScore":35.8,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":34.4,"rawScore":34.4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":35.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":35.3,"upper":36.3},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":65.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":63,"upper":67.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"defaultBaseEdition":"from-20260929","reasoning":{"singleStep":{"score":50},"multiStep":{"score":28},"labels":["builds-on-plan"]}},{"id":"owl-alpha","identityStatus":"revealed","key":"owl-alpha","name":"Owl Alpha","revealedName":"LongCat-2.0","aliases":["openrouter/owl-alpha","LongCat-2.0","LongCat-2.0-Preview","meituan/longcat-2.0"],"group":"Revealed models","url":"https://stealthmodels.com/owl-alpha","route":"meituan/longcat-2.0","providerUrl":"https://openrouter.ai/meituan/longcat-2.0","license":{"label":"MIT","url":"https://huggingface.co/meituan-longcat/LongCat-2.0/blob/main/LICENSE","access":"open"},"artwork":"owl-alpha-card","favicon":"/_shared/favicons/owl-alpha.png","openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":25.6,"rawScore":25.6,"count":7,"countedResults":6,"prompts":6,"bySubject":{"selfie":{"score":61.6,"rawScore":61.6,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":4.1,"rawScore":4.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":54.4,"rawScore":54.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":23.1,"rawScore":23.1,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":12.9,"rawScore":12.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":43.8,"rawScore":43.8,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":79,"rawScore":79,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"latch-five-levels":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0}},"capabilities":{"general":{"score":17.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":8.3,"upper":58.7},"availableFamilies":6,"expectedFamilies":8,"missingFamilies":["foresight-self-improvement","latch-five-levels"],"partial":true},"visual":{"score":34.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":32.8,"upper":36.9},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":35}}},{"id":"ox-alpha","identityStatus":"revealed","key":"ox-alpha","name":"Ox Alpha","benchmarkSlug":"glm-5.3-flash","benchmarkAliases":["ox-alpha","glm-5-3-flash","stealth/ox-alpha","z-ai/glm-5.3-flash"],"revealedName":"GLM-5.3-Flash","aliases":["ox-alpha","stealth/ox-alpha","GLM-5.3-Flash","GLM 5.3 Flash","glm-5.3-flash","glm-5-3-flash","z-ai/glm-5.3-flash","Z.ai GLM-5.3-Flash"],"group":"Revealed models","url":"https://stealthmodels.com/ox-alpha","route":"z-ai/glm-5.3-flash","providerUrl":"https://openrouter.ai/z-ai/glm-5.3-flash","license":{"label":"MIT","url":"https://huggingface.co/zai-org/GLM-5.3-Flash/blob/main/LICENSE","access":"open"},"artwork":"ox-alpha-card","favicon":"/_shared/favicons/ox-alpha.png","openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":48.9,"rawScore":48.9,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":76.6,"rawScore":76.6,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":49.5,"rawScore":49.5,"count":1,"countedResults":1,"prompts":1},"beach":{"score":77.7,"rawScore":77.7,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":27.6,"rawScore":27.6,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":63.8,"rawScore":63.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":56.7,"rawScore":56.7,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":75.8,"rawScore":75.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":49,"rawScore":49,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":4,"rawScore":4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":42.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":41.8,"upper":42.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":59.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":57.6,"upper":61.8},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":43},"multiStep":{"score":22}}},{"id":"union-alpha","identityStatus":"revealed","routerStatus":"confirmed","key":"union-alpha","name":"Union Alpha","revealedName":"Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)","aliases":["Pareto","Pareto 26.9"],"group":"Revealed models","url":"https://union-alpha.com/","route":"unbiased/pareto","providerUrl":"https://openrouter.ai/unbiased/pareto","license":{"label":"Proprietary hosted service","url":"https://unbiased.ai/terms/","access":"proprietary"},"artwork":"union-alpha","favicon":"/_shared/favicons/union-alpha.png","openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":62.7,"rawScore":62.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":76.5,"rawScore":76.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":59.1,"rawScore":59.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":71.5,"rawScore":71.5,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":52,"rawScore":52,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":63.3,"rawScore":63.3,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":70.2,"rawScore":70.2,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":67,"rawScore":67,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":78.7,"rawScore":78.7,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":36.5,"rawScore":36.5,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":58.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":58.2,"upper":59.2},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":69.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":67.8,"upper":71.8},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}}},{"id":"deepseek/deepseek-v4-pro-0813","key":"deepseek-deepseek-v4-pro-0813","name":"DeepSeek-V4-Pro-0813","aliases":[],"group":"Reference models","url":"","route":"deepseek/deepseek-v4-pro-0813","providerUrl":"https://openrouter.ai/deepseek/deepseek-v4-pro-0813","license":{"label":"MIT","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"DeepSeek","logo":"/benchmarks/brands/deepseek.svg"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":42,"rawScore":42,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":42.4,"rawScore":42.4,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":16.9,"rawScore":16.9,"count":1,"countedResults":1,"prompts":1},"beach":{"score":48.2,"rawScore":48.2,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":36.9,"rawScore":36.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":39.6,"rawScore":39.6,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":54,"rawScore":54,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":62.5,"rawScore":62.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":42.7,"rawScore":42.7,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":71.7,"rawScore":71.7,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":41.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":41.1,"upper":42},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":42.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":40.8,"upper":45},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":39},"multiStep":{"score":40}}},{"id":"deepseek/deepseek-v4.1-flash","key":"deepseek-deepseek-v4-1-flash","name":"DeepSeek-V4.1-Flash","aliases":[],"group":"Reference models","url":"","route":"deepseek/deepseek-v4.1-flash","providerUrl":"https://openrouter.ai/deepseek/deepseek-v4.1-flash","license":{"label":"MIT","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"DeepSeek","logo":"/benchmarks/brands/deepseek.svg"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":49.8,"rawScore":49.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":67.9,"rawScore":67.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":37.5,"rawScore":37.5,"count":1,"countedResults":1,"prompts":1},"beach":{"score":59.8,"rawScore":59.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":46.2,"rawScore":46.2,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":39.8,"rawScore":39.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":43.4,"rawScore":43.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":75,"rawScore":75,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":55.4,"rawScore":55.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":65.2,"rawScore":65.2,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":48.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":48,"upper":48.9},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":52.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":50.3,"upper":54.5},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":36},"multiStep":{"score":30}}},{"id":"openai/gpt-6.1-sol","key":"openai-gpt-6-1-sol","name":"GPT-6.1 Sol","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-6.1-sol","providerUrl":"https://openrouter.ai/openai/gpt-6.1-sol","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":67.8,"rawScore":67.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":76.2,"rawScore":76.2,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":61.4,"rawScore":61.4,"count":1,"countedResults":1,"prompts":1},"beach":{"score":76.3,"rawScore":76.3,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":52.9,"rawScore":52.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":67.3,"rawScore":67.3,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":77.5,"rawScore":77.5,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":79,"rawScore":79,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":62,"rawScore":62,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":87.6,"rawScore":87.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":64.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":64,"upper":64.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":74.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":72,"upper":76},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":93},"multiStep":{"score":83},"labels":["builds-on-plan"]}},{"id":"qwen/qwen3.8-flash","key":"qwen-qwen3-8-flash","name":"Qwen3.8-Flash","aliases":[],"group":"Reference models","url":"","route":"qwen/qwen3.8-flash","providerUrl":"https://openrouter.ai/qwen/qwen3.8-flash","license":{"label":"Qwen Community 1.0","url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"Qwen","logo":"/benchmarks/brands/qwen.png"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":46.6,"rawScore":46.6,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":76.4,"rawScore":76.4,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":51.9,"rawScore":51.9,"count":1,"countedResults":1,"prompts":1},"beach":{"score":77.4,"rawScore":77.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":22.4,"rawScore":22.4,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":37,"rawScore":37,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":62.4,"rawScore":62.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":77.5,"rawScore":77.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":29.7,"rawScore":29.7,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":69.8,"rawScore":69.8,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":38.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":37.9,"upper":38.9},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":60.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":58.2,"upper":62.4},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":46},"multiStep":{"score":46}}},{"id":"google/gemini-3.8-flash","key":"google-gemini-3-8-flash","name":"Gemini 3.8 Flash","aliases":[],"group":"Reference models","url":"","route":"google/gemini-3.8-flash","providerUrl":"https://openrouter.ai/google/gemini-3.8-flash","license":{"label":"Proprietary hosted service","url":"https://ai.google.dev/gemini-api/terms","access":"proprietary"},"artwork":"","brand":{"name":"Google","logo":"/benchmarks/brands/google.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":47.9,"rawScore":47.9,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":59,"rawScore":59,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":29.8,"rawScore":29.8,"count":1,"countedResults":1,"prompts":1},"beach":{"score":75.1,"rawScore":75.1,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":30.3,"rawScore":30.3,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":42.3,"rawScore":42.3,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":68.4,"rawScore":68.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":78.8,"rawScore":78.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":45.4,"rawScore":45.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":72.9,"rawScore":72.9,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":43.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":43.4,"upper":44.3},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":55.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":52.9,"upper":57.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":55},"multiStep":{"score":57}}},{"id":"openai/gpt-5.6-sol","key":"openai-gpt-5-6-sol","name":"GPT-5.6 Sol","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-5.6-sol","providerUrl":"https://openrouter.ai/openai/gpt-5.6-sol","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":54.3,"rawScore":54.3,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":61.4,"rawScore":61.4,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":48,"rawScore":48,"count":1,"countedResults":1,"prompts":1},"beach":{"score":68,"rawScore":68,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":32.1,"rawScore":32.1,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":43.5,"rawScore":43.5,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":62.6,"rawScore":62.6,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":67.3,"rawScore":67.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":62.9,"rawScore":62.9,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":83.6,"rawScore":83.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":51.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":51.5,"upper":52.4},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":58.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":56.7,"upper":61},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":64},"multiStep":{"score":83}}},{"id":"openai/gpt-5.6-terra","key":"openai-gpt-5-6-terra","name":"GPT-5.6 Terra","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-5.6-terra","providerUrl":"https://openrouter.ai/openai/gpt-5.6-terra","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":45.9,"rawScore":45.9,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":62.7,"rawScore":62.7,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":41.1,"rawScore":41.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":56.3,"rawScore":56.3,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":27.3,"rawScore":27.3,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":36.7,"rawScore":36.7,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":63.8,"rawScore":63.8,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":65,"rawScore":65,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":30.1,"rawScore":30.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":80,"rawScore":80,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":41.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":41.5,"upper":42.4},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":53,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":50.9,"upper":55.1},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":60},"multiStep":{"score":68},"labels":["builds-on-plan"]}},{"id":"openai/gpt-6-luna","key":"openai-gpt-6-luna","name":"GPT-6 Luna","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-6-luna","providerUrl":"https://openrouter.ai/openai/gpt-6-luna","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":39.1,"rawScore":39.1,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":45.6,"rawScore":45.6,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":47.4,"rawScore":47.4,"count":1,"countedResults":1,"prompts":1},"beach":{"score":59.1,"rawScore":59.1,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":26.6,"rawScore":26.6,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":30.3,"rawScore":30.3,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":63.2,"rawScore":63.2,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":65.8,"rawScore":65.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":29.8,"rawScore":29.8,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":44,"rawScore":44,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":31.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":30.8,"upper":31.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":52.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":50.1,"upper":54.3},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":53},"multiStep":{"score":61}}},{"id":"openai/gpt-6-sol","key":"openai-gpt-6-sol","name":"GPT-6 Sol","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-6-sol","providerUrl":"https://openrouter.ai/openai/gpt-6-sol","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":51.7,"rawScore":51.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":65.5,"rawScore":65.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":48.7,"rawScore":48.7,"count":1,"countedResults":1,"prompts":1},"beach":{"score":70.1,"rawScore":70.1,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":30,"rawScore":30,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":50.7,"rawScore":50.7,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":69.5,"rawScore":69.5,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":80.3,"rawScore":80.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":30.4,"rawScore":30.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":76.6,"rawScore":76.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":46.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":45.7,"upper":46.6},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":61.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":59.5,"upper":63.7},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":67},"multiStep":{"score":80},"labels":["builds-on-plan"]}},{"id":"x-ai/grok-4.7","key":"x-ai-grok-4-7","name":"Grok 4.7","aliases":[],"group":"Reference models","url":"","route":"x-ai/grok-4.7","providerUrl":"https://openrouter.ai/x-ai/grok-4.7","license":{"label":"Proprietary hosted service","url":"https://x.ai/legal/terms-of-service-enterprise","access":"proprietary"},"artwork":"","brand":{"name":"xAI","logo":"/benchmarks/brands/x-ai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":47.5,"rawScore":47.5,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":46.9,"rawScore":46.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":23,"rawScore":23,"count":1,"countedResults":1,"prompts":1},"beach":{"score":52.2,"rawScore":52.2,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":41,"rawScore":41,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":33.8,"rawScore":33.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":56.3,"rawScore":56.3,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":59,"rawScore":59,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":72.6,"rawScore":72.6,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":65.4,"rawScore":65.4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":48.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":48.1,"upper":49},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":45.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":43.5,"upper":47.7},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":46},"multiStep":{"score":58},"labels":["builds-on-plan"]}},{"id":"minimax/minimax-m3","key":"minimax-minimax-m3","name":"MiniMax M3","aliases":[],"group":"Reference models","url":"","route":"minimax/minimax-m3","providerUrl":"https://openrouter.ai/minimax/minimax-m3","license":{"label":"MiniMax Community License","url":"https://huggingface.co/MiniMaxAI/MiniMax-M3/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"MiniMax","logo":"/benchmarks/brands/minimax.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":34.8,"rawScore":34.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":58.2,"rawScore":58.2,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":19.7,"rawScore":19.7,"count":1,"countedResults":1,"prompts":1},"beach":{"score":58,"rawScore":58,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":29.7,"rawScore":29.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":28.8,"rawScore":28.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":66.2,"rawScore":66.2,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":66,"rawScore":66,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":22.1,"rawScore":22.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":31.4,"rawScore":31.4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":28,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":27.5,"upper":28.6},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":45.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":43.5,"upper":47.8},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":46},"multiStep":{"score":20},"labels":["builds-on-plan"]}},{"id":"anthropic/claude-sonnet-5.5","key":"anthropic-claude-sonnet-5-5","name":"Claude Sonnet 5.5","aliases":[],"group":"Reference models","url":"","route":"anthropic/claude-sonnet-5.5","providerUrl":"https://openrouter.ai/anthropic/claude-sonnet-5.5","license":{"label":"Proprietary hosted service","url":"https://www.anthropic.com/legal/commercial-terms","access":"proprietary"},"artwork":"","brand":{"name":"Anthropic","logo":"/benchmarks/brands/anthropic.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":70.3,"rawScore":70.3,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":79.8,"rawScore":79.8,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":60.2,"rawScore":60.2,"count":1,"countedResults":1,"prompts":1},"beach":{"score":87,"rawScore":87,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":52.9,"rawScore":52.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":79.5,"rawScore":79.5,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":76.1,"rawScore":76.1,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":76.8,"rawScore":76.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":59.6,"rawScore":59.6,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":83.9,"rawScore":83.9,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":67,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":66.6,"upper":67.5},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":76.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":74.3,"upper":78.3},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":63},"multiStep":{"score":72},"labels":["builds-on-plan"]}},{"id":"xiaomi/mimo-v2.6-flash","key":"xiaomi-mimo-v2-6-flash","name":"MiMo-V2.6-Flash","aliases":[],"group":"Reference models","url":"","route":"xiaomi/mimo-v2.6-flash","providerUrl":"https://openrouter.ai/xiaomi/mimo-v2.6-flash","license":{"label":"MIT","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL","access":"open"},"artwork":"","brand":{"name":"Xiaomi","logo":"/benchmarks/brands/xiaomi.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":44.7,"rawScore":44.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":67,"rawScore":67,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":32.5,"rawScore":32.5,"count":1,"countedResults":1,"prompts":1},"beach":{"score":67,"rawScore":67,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":12,"rawScore":12,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":50,"rawScore":50,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":51.3,"rawScore":51.3,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":66.5,"rawScore":66.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":71.3,"rawScore":71.3,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":66.3,"rawScore":66.3,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":41,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":40.4,"upper":41.6},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":50.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":48,"upper":52.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":41},"multiStep":{"score":33}}},{"id":"qwen/qwen3.8-27B","key":"qwen-qwen3-8-27b","name":"Qwen3.8-27B","aliases":[],"group":"Reference models","url":"","route":"qwen/qwen3.8-27B","providerUrl":"https://openrouter.ai/qwen/qwen3.8-27B","license":{"label":"Apache-2.0","url":"https://huggingface.co/Qwen/Qwen3.8-27B","access":"open"},"artwork":"","brand":{"name":"Qwen","logo":"/benchmarks/brands/qwen.png"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":51.7,"rawScore":51.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":77.5,"rawScore":77.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":47.2,"rawScore":47.2,"count":1,"countedResults":1,"prompts":1},"beach":{"score":68.9,"rawScore":68.9,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":40.2,"rawScore":40.2,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":42.1,"rawScore":42.1,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":67.7,"rawScore":67.7,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":70.5,"rawScore":70.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":63,"rawScore":63,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":38.6,"rawScore":38.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":46.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":45.7,"upper":46.7},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":61.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":59.3,"upper":63.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":39},"multiStep":{"score":50}}},{"id":"anthropic/claude-fable-5.1","key":"anthropic-claude-fable-5-1","name":"Claude Fable 5.1","aliases":[],"group":"Reference models","url":"","route":"anthropic/claude-fable-5.1","providerUrl":"https://openrouter.ai/anthropic/claude-fable-5.1","license":{"label":"Proprietary hosted service","url":"https://www.anthropic.com/legal/commercial-terms","access":"proprietary"},"artwork":"","brand":{"name":"Anthropic","logo":"/benchmarks/brands/anthropic.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":62.1,"rawScore":62.1,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":68.2,"rawScore":68.2,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":57.6,"rawScore":57.6,"count":1,"countedResults":1,"prompts":1},"beach":{"score":73.8,"rawScore":73.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":54.7,"rawScore":54.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":64.1,"rawScore":64.1,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":70,"rawScore":70,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":72.5,"rawScore":72.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":34.2,"rawScore":34.2,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":80.2,"rawScore":80.2,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":58.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":58.4,"upper":59.4},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":67.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":65.1,"upper":69.4},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":72},"multiStep":{"score":84}}},{"id":"anthropic/claude-opus-5.5","key":"anthropic-claude-opus-5-5","name":"Claude Opus 5.5","aliases":[],"group":"Reference models","url":"","route":"anthropic/claude-opus-5.5","providerUrl":"https://openrouter.ai/anthropic/claude-opus-5.5","license":{"label":"Proprietary hosted service","url":"https://www.anthropic.com/legal/commercial-terms","access":"proprietary"},"artwork":"","brand":{"name":"Anthropic","logo":"/benchmarks/brands/anthropic.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":72.7,"rawScore":72.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":75.2,"rawScore":75.2,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":62.5,"rawScore":62.5,"count":1,"countedResults":1,"prompts":1},"beach":{"score":77,"rawScore":77,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":61.1,"rawScore":61.1,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":80.6,"rawScore":80.6,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":76,"rawScore":76,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":76,"rawScore":76,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":66.3,"rawScore":66.3,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":87.6,"rawScore":87.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":71.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":70.9,"upper":71.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":75.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":73,"upper":77.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":78},"multiStep":{"score":72},"labels":["builds-on-plan"]}},{"id":"z-ai/glm-5.3","key":"z-ai-glm-5-3","name":"GLM-5.3","aliases":[],"group":"Reference models","url":"","route":"z-ai/glm-5.3","providerUrl":"https://openrouter.ai/z-ai/glm-5.3","license":{"label":"GLM-5.3 License","url":"https://huggingface.co/zai-org/GLM-5.3/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"Z.AI","logo":"/benchmarks/brands/z-ai.svg"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":45.4,"rawScore":45.4,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":76.5,"rawScore":76.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":30,"rawScore":30,"count":1,"countedResults":1,"prompts":1},"beach":{"score":75.7,"rawScore":75.7,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":25.4,"rawScore":25.4,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":49.8,"rawScore":49.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":54.7,"rawScore":54.7,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":56.5,"rawScore":56.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":59.1,"rawScore":59.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":25.6,"rawScore":25.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":37.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":37.2,"upper":38.2},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":58,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":55.9,"upper":60.1},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":62},"multiStep":{"score":34}}},{"id":"openai/gpt-6-astra","key":"openai-gpt-6-astra","name":"GPT-6 Astra","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-6-astra","providerUrl":"https://openrouter.ai/openai/gpt-6-astra","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":74.7,"rawScore":74.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":80.9,"rawScore":80.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":66.9,"rawScore":66.9,"count":1,"countedResults":1,"prompts":1},"beach":{"score":80.6,"rawScore":80.6,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":70.6,"rawScore":70.6,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":69.2,"rawScore":69.2,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":78,"rawScore":78,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":76,"rawScore":76,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":76.1,"rawScore":76.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":87.6,"rawScore":87.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":73.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":73,"upper":73.9},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":77.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":75.1,"upper":79.1},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":88},"multiStep":{"score":87},"labels":["builds-on-plan"]}},{"id":"moonshotai/kimi-k3","key":"moonshotai-kimi-k3","name":"Kimi K3","aliases":[],"group":"Reference models","url":"","route":"moonshotai/kimi-k3","providerUrl":"https://openrouter.ai/moonshotai/kimi-k3","license":{"label":"Kimi K3 License","url":"https://huggingface.co/moonshotai/Kimi-K3/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"Moonshot AI","logo":"/benchmarks/brands/moonshot.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":50.7,"rawScore":50.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":78.8,"rawScore":78.8,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":42.4,"rawScore":42.4,"count":1,"countedResults":1,"prompts":1},"beach":{"score":75.6,"rawScore":75.6,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":35.3,"rawScore":35.3,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":40,"rawScore":40,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":71.8,"rawScore":71.8,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":66.3,"rawScore":66.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":52.7,"rawScore":52.7,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":57.7,"rawScore":57.7,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":44.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":44,"upper":45},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":61.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":59.6,"upper":63.8},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":48},"multiStep":{"score":71}}},{"id":"meta/muse-spark-1.3","key":"meta-muse-spark-1-3","name":"Muse Spark 1.3","aliases":[],"group":"Reference models","url":"","route":"meta/muse-spark-1.3","providerUrl":"https://openrouter.ai/meta/muse-spark-1.3","license":{"label":"Proprietary hosted service","url":"https://dev.meta.ai/legal/terms-of-service","access":"proprietary"},"artwork":"","brand":{"name":"Meta","logo":"/benchmarks/brands/meta.svg"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":39.8,"rawScore":39.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":63.4,"rawScore":63.4,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":15,"rawScore":15,"count":1,"countedResults":1,"prompts":1},"beach":{"score":44.8,"rawScore":44.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":16.5,"rawScore":16.5,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":28.6,"rawScore":28.6,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":66,"rawScore":66,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":66.8,"rawScore":66.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":64.3,"rawScore":64.3,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":72.4,"rawScore":72.4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":38.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":37.8,"upper":38.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":42.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":40.3,"upper":44.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":35},"multiStep":{"score":68}}},{"id":"openai/gpt-5.5","key":"openai-gpt-5-5","name":"GPT-5.5","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-5.5","providerUrl":"https://openrouter.ai/openai/gpt-5.5","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":31.5,"rawScore":31.5,"count":7,"countedResults":6,"prompts":6,"bySubject":{"selfie":{"score":41.9,"rawScore":41.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":17.3,"rawScore":17.3,"count":1,"countedResults":1,"prompts":1},"beach":{"score":34.9,"rawScore":34.9,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":22.7,"rawScore":22.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":40.8,"rawScore":40.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":37.4,"rawScore":37.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":60,"rawScore":60,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"latch-five-levels":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0}},"capabilities":{"general":{"score":26.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":13.1,"upper":63.5},"availableFamilies":6,"expectedFamilies":8,"missingFamilies":["foresight-self-improvement","latch-five-levels"],"partial":true},"visual":{"score":36.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":34.5,"upper":38.8},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":34}}},{"id":"openai/gpt-5.6-luna","key":"openai-gpt-5-6-luna","name":"GPT-5.6 Luna","aliases":[],"group":"Reference models","url":"","route":"openai/gpt-5.6-luna","providerUrl":"https://openrouter.ai/openai/gpt-5.6-luna","license":{"label":"Proprietary hosted service","url":"https://openai.com/policies/services-agreement/","access":"proprietary"},"artwork":"","brand":{"name":"OpenAI","logo":"/benchmarks/brands/openai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":37.9,"rawScore":37.9,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":59.8,"rawScore":59.8,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":39.1,"rawScore":39.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":47.8,"rawScore":47.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":21.9,"rawScore":21.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":32.7,"rawScore":32.7,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":58.4,"rawScore":58.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":66.3,"rawScore":66.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":25.1,"rawScore":25.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":52.9,"rawScore":52.9,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":30.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":29.9,"upper":30.9},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":50.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":48.4,"upper":52.7},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":33},"multiStep":{"score":60},"labels":["builds-on-plan"]}},{"id":"fledge-alpha","identityStatus":"stealth","routerStatus":"suspected","key":"fledge-alpha","name":"Fledge Alpha","aliases":["fledge-alpha","fledge-alpha-free","fledge alpha","opencode/fledge-alpha-free"],"group":"Stealth models","url":"https://stealthmodels.com/fledge-alpha","route":"opencode/fledge-alpha-free","providerUrl":"https://opencode.ai/zen","license":{"label":"Not disclosed","access":"proprietary"},"artwork":"fledge-alpha","favicon":"/_shared/favicons/fledge-alpha.png","openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":36,"rawScore":36,"count":7,"countedResults":6,"prompts":6,"bySubject":{"selfie":{"score":61.1,"rawScore":61.1,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":28.3,"rawScore":28.3,"count":1,"countedResults":1,"prompts":1},"beach":{"score":43.4,"rawScore":43.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":16.4,"rawScore":16.4,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":36.8,"rawScore":36.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":65.8,"rawScore":65.8,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":60.5,"rawScore":60.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"latch-five-levels":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0}},"capabilities":{"general":{"score":27.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":13.3,"upper":63.7},"availableFamilies":6,"expectedFamilies":8,"missingFamilies":["foresight-self-improvement","latch-five-levels"],"partial":true},"visual":{"score":45.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":43.5,"upper":47.7},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":44}}},{"id":"tencent/hy4-preview","key":"tencent-hy4-preview","name":"Hy4 preview","aliases":[],"group":"Reference models","url":"","route":"tencent/hy4-preview","providerUrl":"https://openrouter.ai/tencent/hy4-preview","license":{"label":"Apache-2.0","url":"https://github.com/Tencent-Hunyuan/Hy4-preview/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"Tencent Hunyuan","logo":"/benchmarks/brands/tencent.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":44.1,"rawScore":44.1,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":60.2,"rawScore":60.2,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":23.4,"rawScore":23.4,"count":1,"countedResults":1,"prompts":1},"beach":{"score":60,"rawScore":60,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":41.4,"rawScore":41.4,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":41.3,"rawScore":41.3,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":55.5,"rawScore":55.5,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":60,"rawScore":60,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":53.1,"rawScore":53.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":35.6,"rawScore":35.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":41.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":40.8,"upper":41.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":49.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":47,"upper":51.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":30},"multiStep":{"score":23},"labels":["re-reasons"]}},{"id":"qwen/qwen3.8-max-0902","key":"qwen-qwen3-8-max-0902","name":"Qwen3.8-Max-0902","aliases":[],"group":"Reference models","url":"","route":"qwen/qwen3.8-max-0902","providerUrl":"https://openrouter.ai/qwen/qwen3.8-max-0902","license":{"label":"Proprietary hosted service","url":"https://www.alibabacloud.com/help/en/model-studio/related-agreements","access":"proprietary"},"artwork":"","brand":{"name":"Qwen","logo":"/benchmarks/brands/qwen.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":63.3,"rawScore":63.3,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":70.6,"rawScore":70.6,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":45.1,"rawScore":45.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":77.7,"rawScore":77.7,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":44.7,"rawScore":44.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":61.8,"rawScore":61.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":69.9,"rawScore":69.9,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":72,"rawScore":72,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":72.4,"rawScore":72.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":82.6,"rawScore":82.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":62.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":62.2,"upper":63.2},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":64.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":62.4,"upper":66.7},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":53},"multiStep":{"score":61}}},{"id":"nvidia/nemotron-3-ultra-550b-a55b","key":"nvidia-nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra","aliases":[],"group":"Reference models","url":"","route":"nvidia/nemotron-3-ultra-550b-a55b","providerUrl":"https://openrouter.ai/nvidia/nemotron-3-ultra-550b-a55b","license":{"label":"OpenMDW License Agreement v1.1","url":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4","access":"open"},"artwork":"","brand":{"name":"NVIDIA","logo":"/benchmarks/brands/nvidia.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":16.8,"rawScore":16.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":36.2,"rawScore":36.2,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":0.5,"rawScore":0.5,"count":1,"countedResults":1,"prompts":1},"beach":{"score":39,"rawScore":39,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":9.9,"rawScore":9.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":20.5,"rawScore":20.5,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":24.3,"rawScore":24.3,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":34.8,"rawScore":34.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":16.9,"rawScore":16.9,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":5,"rawScore":5,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":12.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":11.7,"upper":12.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":26.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":24.7,"upper":28.7},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":29},"multiStep":{"score":32}}},{"id":"Aleph-Alpha/Kolibri-1","key":"aleph-alpha-kolibri-1","name":"Kolibri-1","aliases":[],"group":"Reference models","url":"","route":"Aleph-Alpha/Kolibri-1","providerUrl":"https://huggingface.co/Aleph-Alpha/Kolibri-1","license":{"label":"Apache-2.0","url":"https://huggingface.co/Aleph-Alpha/Kolibri-1","access":"open"},"artwork":"","brand":{"name":"Aleph Alpha","logo":"/benchmarks/brands/aleph-alpha.svg"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":13.6,"rawScore":13.6,"count":7,"countedResults":6,"prompts":6,"bySubject":{"selfie":{"score":22.9,"rawScore":22.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":3.3,"rawScore":3.3,"count":1,"countedResults":1,"prompts":1},"beach":{"score":12.9,"rawScore":12.9,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":10.4,"rawScore":10.4,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":21.2,"rawScore":21.2,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":10,"rawScore":10,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":17.5,"rawScore":17.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"latch-five-levels":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0}},"capabilities":{"general":{"score":12.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":5.9,"upper":56.3},"availableFamilies":6,"expectedFamilies":8,"missingFamilies":["foresight-self-improvement","latch-five-levels"],"partial":true},"visual":{"score":15,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":13.1,"upper":17.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":7}}},{"id":"qwen/qwen3.6-35b-a3b","key":"qwen-qwen3-6-35b-a3b","name":"Qwen3.6-35B-A3B","aliases":[],"group":"Reference models","url":"","route":"qwen/qwen3.6-35b-a3b","providerUrl":"https://openrouter.ai/qwen/qwen3.6-35b-a3b","license":{"label":"Apache-2.0","url":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B","access":"open"},"artwork":"","brand":{"name":"Qwen","logo":"/benchmarks/brands/qwen.png"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":21.1,"rawScore":21.1,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":63.5,"rawScore":63.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":13.8,"rawScore":13.8,"count":1,"countedResults":1,"prompts":1},"beach":{"score":50.7,"rawScore":50.7,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":9.1,"rawScore":9.1,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":12.8,"rawScore":12.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":43.6,"rawScore":43.6,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":53.8,"rawScore":53.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":20.7,"rawScore":20.7,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":3.6,"rawScore":3.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":12.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":12.2,"upper":13.2},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":34.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":32.7,"upper":37},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":42},"multiStep":{"score":38}}},{"id":"qwen/qwen3.5-9b","key":"qwen-qwen3-5-9b","name":"Qwen3.5-9B","aliases":[],"group":"Reference models","url":"","route":"qwen/qwen3.5-9b","providerUrl":"https://openrouter.ai/qwen/qwen3.5-9b","license":{"label":"Apache-2.0","url":"https://huggingface.co/Qwen/Qwen3.5-9B","access":"open"},"artwork":"","brand":{"name":"Qwen","logo":"/benchmarks/brands/qwen.png"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":10.5,"rawScore":10.5,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":25.8,"rawScore":25.8,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":1.2,"rawScore":1.2,"count":1,"countedResults":1,"prompts":1},"beach":{"score":12.8,"rawScore":12.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":9.7,"rawScore":9.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":12,"rawScore":12,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":14.7,"rawScore":14.7,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":23,"rawScore":23,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":12.8,"rawScore":12.8,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":3.8,"rawScore":3.8,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":7.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":7.4,"upper":8.3},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":14.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":13.3,"upper":17},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":14},"multiStep":{"score":16}}},{"id":"inclusionai/ling-3.1-flash","key":"inclusionai-ling-3-1-flash","name":"Ling 3.1 Flash","aliases":[],"group":"Reference models","url":"","route":"inclusionai/ling-3.1-flash","providerUrl":"https://openrouter.ai/inclusionai/ling-3.1-flash","license":{"label":"MIT","url":"https://huggingface.co/inclusionAI/Ling-3.1-flash","access":"open"},"artwork":"","brand":{"name":"InclusionAI","logo":"/benchmarks/brands/inclusionai.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":42.2,"rawScore":42.2,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":73.2,"rawScore":73.2,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":24,"rawScore":24,"count":1,"countedResults":1,"prompts":1},"beach":{"score":65.7,"rawScore":65.7,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":37.9,"rawScore":37.9,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":30.6,"rawScore":30.6,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":63.1,"rawScore":63.1,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":64.3,"rawScore":64.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":43.1,"rawScore":43.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":45.9,"rawScore":45.9,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":35.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":34.9,"upper":36.2},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":51.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":49.1,"upper":53.3},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":39},"multiStep":{"score":31}}},{"id":"mistralai/mistral-large-4-0","key":"mistralai-mistral-large-4-0","name":"Mistral Large 4","aliases":[],"group":"Reference models","url":"","route":"mistralai/mistral-large-4-0","providerUrl":"https://openrouter.ai/mistralai/mistral-large-4-0","license":{"label":"Hosted service; weights license not established","url":"https://legal.mistral.ai/terms/commercial-terms-of-service","access":"proprietary"},"artwork":"","brand":{"name":"Mistral AI","logo":"/benchmarks/brands/mistral.svg"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":43.3,"rawScore":43.3,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":72.5,"rawScore":72.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":41.7,"rawScore":41.7,"count":1,"countedResults":1,"prompts":1},"beach":{"score":65.4,"rawScore":65.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":28.8,"rawScore":28.8,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":38.1,"rawScore":38.1,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":69.5,"rawScore":69.5,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":70.5,"rawScore":70.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":22,"rawScore":22,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":49.4,"rawScore":49.4,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":34.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":33.9,"upper":34.9},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":58,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":55.8,"upper":60.1},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":46},"multiStep":{"score":38}}},{"id":"anthropic/claude-haiku-5.5","key":"anthropic-claude-haiku-5-5","name":"Claude Haiku 5.5","aliases":[],"group":"Reference models","url":"","route":"anthropic/claude-haiku-5.5","providerUrl":"https://openrouter.ai/anthropic/claude-haiku-5.5","license":{"label":"Proprietary hosted service","url":"https://www.anthropic.com/legal/commercial-terms","access":"proprietary"},"artwork":"","brand":{"name":"Anthropic","logo":"/benchmarks/brands/anthropic.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":49.8,"rawScore":49.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":46.3,"rawScore":46.3,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":38.6,"rawScore":38.6,"count":1,"countedResults":1,"prompts":1},"beach":{"score":70.3,"rawScore":70.3,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":40.7,"rawScore":40.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":49.2,"rawScore":49.2,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":47.3,"rawScore":47.3,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":61.8,"rawScore":61.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":34.4,"rawScore":34.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":78.3,"rawScore":78.3,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":47.7,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":47.2,"upper":48.1},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":53.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":51.4,"upper":55.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":38},"multiStep":{"score":55},"labels":["builds-on-plan"]}},{"id":"glyph-cluster","identityStatus":"stealth","key":"glyph-cluster","name":"Glyph Cluster","aliases":["stealth/glyph-cluster"],"group":"Stealth models","url":"https://stealthmodels.com/glyph-cluster","route":"stealth/glyph-cluster","providerUrl":"https://vercel.com/ai-gateway/models/glyph-cluster","license":{"label":"Hosted preview","url":"https://vercel.com/changelog/glyph-cluster-is-now-available-in-stealth-for-free-on-ai-gateway","access":"proprietary"},"artwork":"glyph-cluster","favicon":"/_shared/favicons/glyph-cluster.png","openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":19.5,"rawScore":19.5,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":32.4,"rawScore":32.4,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":11.7,"rawScore":11.7,"count":1,"countedResults":1,"prompts":1},"beach":{"score":46.3,"rawScore":46.3,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":12.8,"rawScore":12.8,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":19.9,"rawScore":19.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":42.8,"rawScore":42.8,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":44.3,"rawScore":44.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":6.4,"rawScore":6.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":3.6,"rawScore":3.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":12.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":11.6,"upper":12.7},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":31.1,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":29,"upper":33.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":72},"multiStep":{"score":41}}},{"id":"qwen/qwen3.6-27b","key":"qwen-qwen3-6-27b","name":"Qwen3.6-27B","aliases":[],"group":"Reference models","url":"","route":"qwen/qwen3.6-27b","providerUrl":"https://openrouter.ai/qwen/qwen3.6-27b","license":{"label":"Apache-2.0","url":"https://huggingface.co/Qwen/Qwen3.6-27B/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"Qwen","logo":"/benchmarks/brands/qwen.png"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":32.5,"rawScore":32.5,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":56.8,"rawScore":56.8,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":30.8,"rawScore":30.8,"count":1,"countedResults":1,"prompts":1},"beach":{"score":56.7,"rawScore":56.7,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":15.5,"rawScore":15.5,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":26.7,"rawScore":26.7,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":45.4,"rawScore":45.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":58.5,"rawScore":58.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":37.8,"rawScore":37.8,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":28.5,"rawScore":28.5,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":25.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":25.1,"upper":26.2},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":44.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":42.1,"upper":46.3},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":39},"multiStep":{"score":36}}},{"id":"microsoft/mai-thinking-1","key":"microsoft-mai-thinking-1","name":"MAI-Thinking-1","aliases":[],"group":"Reference models","url":"","route":"microsoft/mai-thinking-1","providerUrl":"https://openrouter.ai/microsoft/mai-thinking-1","license":{"label":"Proprietary hosted service","access":"proprietary"},"artwork":"","brand":{"name":"Microsoft AI","logo":"/benchmarks/brands/microsoft-ai.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":14.6,"rawScore":14.6,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":46.3,"rawScore":46.3,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":11.8,"rawScore":11.8,"count":1,"countedResults":1,"prompts":1},"beach":{"score":44.2,"rawScore":44.2,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":8,"rawScore":8,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":1.9,"rawScore":1.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":37.3,"rawScore":37.3,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":56.5,"rawScore":56.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":6.4,"rawScore":6.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":3.6,"rawScore":3.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":7.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":7,"upper":8.1},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":25.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":23.6,"upper":27.9},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"multiStep":{"score":58}}},{"id":"google/gemma-4-31b-it","key":"google-gemma-4-31b-it","name":"Gemma 4 31B","aliases":[],"group":"Reference models","url":"","route":"google/gemma-4-31b-it","providerUrl":"https://openrouter.ai/google/gemma-4-31b-it","license":{"label":"Apache 2.0","url":"https://ai.google.dev/gemma/apache_2","access":"open"},"artwork":"","brand":{"name":"Google","logo":"/benchmarks/brands/google.png"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":26.6,"rawScore":26.6,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":60.9,"rawScore":60.9,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":35.1,"rawScore":35.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":29.2,"rawScore":29.2,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":16,"rawScore":16,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":14.1,"rawScore":14.1,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":38.2,"rawScore":38.2,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":37.8,"rawScore":37.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":30.2,"rawScore":30.2,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":26.6,"rawScore":26.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":19.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":18.8,"upper":19.8},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":38.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":36.4,"upper":40.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":87},"multiStep":{"score":48}}},{"id":"google/gemma-4-26b-a4b-it","key":"google-gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B","aliases":[],"group":"Reference models","url":"","route":"google/gemma-4-26b-a4b-it","providerUrl":"https://openrouter.ai/google/gemma-4-26b-a4b-it","license":{"label":"Apache 2.0","url":"https://ai.google.dev/gemma/apache_2","access":"open"},"artwork":"","brand":{"name":"Google","logo":"/benchmarks/brands/google.png"},"openModel":true,"localCapable":true,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":17.2,"rawScore":17.2,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":52.8,"rawScore":52.8,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":17.8,"rawScore":17.8,"count":1,"countedResults":1,"prompts":1},"beach":{"score":28.5,"rawScore":28.5,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":10,"rawScore":10,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":14.2,"rawScore":14.2,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":40.3,"rawScore":40.3,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":34,"rawScore":34,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":4.1,"rawScore":4.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":4.5,"rawScore":4.5,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":9.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":9.1,"upper":10.1},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":30.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":28.3,"upper":32.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":84},"multiStep":{"score":16},"labels":["re-reasons"]}},{"id":"missingno","identityStatus":"stealth","key":"missingno","name":"MISSINGNO.","aliases":["MissingNo","MissingNo.","missing-no","missing no","stealth/missingno"],"group":"Stealth models","url":"https://stealthmodels.com/missingno","route":"missingno","providerUrl":"","license":{"label":"Not disclosed","access":"proprietary"},"artwork":"","favicon":"/_shared/favicons/missingno.svg","openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":38.8,"rawScore":38.8,"count":7,"countedResults":6,"prompts":6,"bySubject":{"selfie":{"score":49.7,"rawScore":49.7,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":28.1,"rawScore":28.1,"count":1,"countedResults":1,"prompts":1},"beach":{"score":43.9,"rawScore":43.9,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":27.8,"rawScore":27.8,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":42.8,"rawScore":42.8,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":57.5,"rawScore":57.5,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":58.8,"rawScore":58.8,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0},"latch-five-levels":{"score":null,"rawScore":null,"count":0,"countedResults":0,"prompts":0}},"capabilities":{"general":{"score":35.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":17.4,"upper":67.8},"availableFamilies":6,"expectedFamilies":8,"missingFamilies":["foresight-self-improvement","latch-five-levels"],"partial":true},"visual":{"score":42.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":40.5,"upper":44.8},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":32}}},{"id":"meta/muse-wonder","key":"meta-muse-wonder","name":"Muse Wonder","aliases":[],"group":"Reference models","url":"","route":"meta/muse-wonder","providerUrl":"https://openrouter.ai/meta/muse-wonder","license":{"label":"Proprietary hosted service","url":"https://dev.meta.ai/legal/terms-of-service","access":"proprietary"},"artwork":"","brand":{"name":"Meta","logo":"/benchmarks/brands/meta.svg"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":57.3,"rawScore":57.3,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":71.5,"rawScore":71.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":55.7,"rawScore":55.7,"count":1,"countedResults":1,"prompts":1},"beach":{"score":71,"rawScore":71,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":39.7,"rawScore":39.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":59.9,"rawScore":59.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":66.5,"rawScore":66.5,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":72.3,"rawScore":72.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":52.3,"rawScore":52.3,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":67,"rawScore":67,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":53.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":52.7,"upper":53.7},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":64.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":62.8,"upper":67},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":59},"multiStep":{"score":80}}},{"id":"google/gemini-4-argon","key":"google-gemini-4-argon","name":"Gemini 4 Argon","aliases":[],"group":"Reference models","url":"","route":"google/gemini-4-argon","providerUrl":"https://openrouter.ai/google/gemini-4-argon","license":{"label":"Proprietary hosted service","url":"https://ai.google.dev/gemini-api/terms","access":"proprietary"},"artwork":"","brand":{"name":"Google","logo":"/benchmarks/brands/google.png"},"openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":67.3,"rawScore":67.3,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":71.6,"rawScore":71.6,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":65.7,"rawScore":65.7,"count":1,"countedResults":1,"prompts":1},"beach":{"score":82.5,"rawScore":82.5,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":58,"rawScore":58,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":68.4,"rawScore":68.4,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":80.6,"rawScore":80.6,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":84.5,"rawScore":84.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":48.6,"rawScore":48.6,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":82.6,"rawScore":82.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":63.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":63.2,"upper":64.1},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":74.3,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":72.2,"upper":76.4},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":81},"multiStep":{"score":75}}},{"id":"cat-jam","identityStatus":"revealed","key":"cat-jam","name":"Cat Jam Free","aliases":["Cat Jam","Cat Jam Free","cat-jam-free","opencode/cat-jam-free","opencode-go/cat-jam-free","stealth/cat-jam"],"group":"Revealed models","url":"https://stealthmodels.com/cat-jam","route":"stealth/cat-jam","providerUrl":"https://opencode.ai/v2/docs/console/models/","license":{"label":"Apache-2.0","url":"https://huggingface.co/reflection/Beam-1.0-501B-A23B/blob/main/LICENSE","access":"proprietary"},"artwork":"","favicon":"/_shared/favicons/cat-jam.png","openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":33.9,"rawScore":33.9,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":54.6,"rawScore":54.6,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":30.3,"rawScore":30.3,"count":1,"countedResults":1,"prompts":1},"beach":{"score":48.1,"rawScore":48.1,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":17.2,"rawScore":17.2,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":30.9,"rawScore":30.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":57.1,"rawScore":57.1,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":58,"rawScore":58,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":33.8,"rawScore":33.8,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":33.9,"rawScore":33.9,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":27.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":27.3,"upper":28.3},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":43.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":41.8,"upper":46},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":41},"multiStep":{"score":32}}},{"id":"meta/muse-spark-open","key":"meta-muse-spark-open","name":"Muse Spark Open","aliases":[],"group":"Reference models","url":"","route":"meta/muse-spark-open","providerUrl":"https://vercel.com/ai-gateway/models/muse-spark-open","license":{"label":"Muse Model License","url":"https://huggingface.co/unsloth/Muse-Spark-Open-816B-A42B-GGUF/blob/main/LICENSE.md","access":"open"},"artwork":"","brand":{"name":"Meta","logo":"/benchmarks/brands/meta.svg"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":44.4,"rawScore":44.4,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":52.1,"rawScore":52.1,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":37.8,"rawScore":37.8,"count":1,"countedResults":1,"prompts":1},"beach":{"score":64.2,"rawScore":64.2,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":19.6,"rawScore":19.6,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":51.9,"rawScore":51.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":48.1,"rawScore":48.1,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":60.5,"rawScore":60.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":29.4,"rawScore":29.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":75.7,"rawScore":75.7,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":42.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":42.1,"upper":43.1},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":47.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":45.4,"upper":49.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":36},"multiStep":{"score":53},"labels":["builds-on-plan"]}},{"id":"reflection/beam-1.0-501b-a23b","key":"reflection-beam-1-0-501b-a23b","name":"Beam 1.0 501B A23B","aliases":[],"group":"Reference models","url":"","route":"reflection/beam-1.0-501b-a23b","providerUrl":"https://openrouter.ai/reflection/beam-1.0-501b-a23b","license":{"label":"Apache 2.0","url":"https://huggingface.co/reflection/Beam-1.0-501B-A23B/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"Reflection AI","logo":"/benchmarks/brands/reflection-ai.svg"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":32.6,"rawScore":32.6,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":45.4,"rawScore":45.4,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":0.7,"rawScore":0.7,"count":1,"countedResults":1,"prompts":1},"beach":{"score":52.4,"rawScore":52.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":21.7,"rawScore":21.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":27.9,"rawScore":27.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":55.2,"rawScore":55.2,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":46.3,"rawScore":46.3,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":49.4,"rawScore":49.4,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":37.3,"rawScore":37.3,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":31,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":30.6,"upper":31.5},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":35.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":34,"upper":37.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":30},"multiStep":{"score":32}}},{"id":"deepseek/deepseek-v4.1-flash-1017","key":"deepseek-deepseek-v4-1-flash-1017","name":"DeepSeek-V4.1-Flash-1017","aliases":[],"group":"Reference models","url":"","route":"deepseek/deepseek-v4.1-flash-1017","providerUrl":"https://openrouter.ai/deepseek/deepseek-v4.1-flash-1017","license":{"label":"MIT","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"DeepSeek","logo":"/benchmarks/brands/deepseek.svg"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":59.2,"rawScore":59.2,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":74.5,"rawScore":74.5,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":54.6,"rawScore":54.6,"count":1,"countedResults":1,"prompts":1},"beach":{"score":57.4,"rawScore":57.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":48.7,"rawScore":48.7,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":60.6,"rawScore":60.6,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":71.4,"rawScore":71.4,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":73.5,"rawScore":73.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":52.9,"rawScore":52.9,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":73.9,"rawScore":73.9,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":56.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":56.4,"upper":57.3},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":63.5,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":61.4,"upper":65.6},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":51},"multiStep":{"score":31}}},{"id":"minimax/minimax-m3.1-flash","key":"minimax-minimax-m3-1-flash","name":"MiniMax M3.1 Flash","aliases":[],"group":"Reference models","url":"","route":"minimax/minimax-m3.1-flash","providerUrl":"https://openrouter.ai/minimax/minimax-m3.1-flash:free","license":{"label":"MiniMax Community License","url":"https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash/blob/main/LICENSE","access":"open"},"artwork":"","brand":{"name":"MiniMax","logo":"/benchmarks/brands/minimax.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":45.9,"rawScore":45.9,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":61.7,"rawScore":61.7,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":50,"rawScore":50,"count":1,"countedResults":1,"prompts":1},"beach":{"score":71.8,"rawScore":71.8,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":39.8,"rawScore":39.8,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":41.9,"rawScore":41.9,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":76.8,"rawScore":76.8,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":72,"rawScore":72,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":30.7,"rawScore":30.7,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":21.9,"rawScore":21.9,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":35.8,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":35.3,"upper":36.4},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":62.6,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":60.5,"upper":64.7},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":50},"multiStep":{"score":40}}},{"id":"xiaomi/mimo-v2.6-pro","key":"xiaomi-mimo-v2-6-pro","name":"MiMo-V2.6-Pro","aliases":[],"group":"Reference models","url":"","route":"xiaomi/mimo-v2.6-pro","providerUrl":"https://openrouter.ai/xiaomi/mimo-v2.6-pro","license":{"label":"MIT","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL","access":"open"},"artwork":"","brand":{"name":"Xiaomi","logo":"/benchmarks/brands/xiaomi.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":48.1,"rawScore":48.1,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":58,"rawScore":58,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":39.9,"rawScore":39.9,"count":1,"countedResults":1,"prompts":1},"beach":{"score":60.9,"rawScore":60.9,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":39,"rawScore":39,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":43.6,"rawScore":43.6,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":56.5,"rawScore":56.5,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":58,"rawScore":58,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":54.6,"rawScore":54.6,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":52.6,"rawScore":52.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":44.4,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":43.9,"upper":44.9},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":55,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":52.8,"upper":57.1},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":35},"multiStep":{"score":39}}},{"id":"moonshotai/kimi-k3.1-pro","key":"moonshotai-kimi-k3-1-pro","name":"Kimi K3.1 Pro","aliases":[],"group":"Reference models","url":"","route":"moonshotai/kimi-k3.1-pro","providerUrl":"https://openrouter.ai/moonshotai/kimi-k3.1-pro","license":{"label":"Not yet published (weights due 4 Nov 2026)","url":"https://www.kimi.com/en/blog/kimi-k3-1-pro","access":"open"},"artwork":"","brand":{"name":"Moonshot AI","logo":"/benchmarks/brands/moonshot.png"},"openModel":true,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":56.7,"rawScore":56.7,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":67.4,"rawScore":67.4,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":55.3,"rawScore":55.3,"count":1,"countedResults":1,"prompts":1},"beach":{"score":64.3,"rawScore":64.3,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":41.1,"rawScore":41.1,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":45.3,"rawScore":45.3,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":61.8,"rawScore":61.8,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":73.5,"rawScore":73.5,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":61.1,"rawScore":61.1,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":81.8,"rawScore":81.8,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":53.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":53.5,"upper":54.4},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":62,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":59.9,"upper":64.2},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":49},"multiStep":{"score":79}}},{"id":"iceman-contributor","identityStatus":"stealth","key":"iceman-contributor","name":"Iceman Contributor","aliases":["Iceman","Iceman Contributor","Iceman Contributor Free","iceman-contributor-free","opencode-go/iceman-contributor-free","opencode/iceman-contributor-free"],"group":"Stealth models","url":"https://stealthmodels.com/iceman-contributor","route":"iceman-contributor","providerUrl":"","license":{"label":"Not disclosed","access":"proprietary"},"artwork":"iceman-contributor-mark","favicon":"/_shared/favicons/iceman-contributor.png","openModel":false,"localCapable":false,"allowPreviousGenerations":false,"defaultEdition":"","scoreSummaries":{"":{"score":19.8,"rawScore":19.8,"count":9,"countedResults":8,"prompts":8,"bySubject":{"selfie":{"score":39.3,"rawScore":39.3,"count":1,"countedResults":1,"prompts":1},"mona-lisa":{"score":19.3,"rawScore":19.3,"count":1,"countedResults":1,"prompts":1},"beach":{"score":46.4,"rawScore":46.4,"count":1,"countedResults":1,"prompts":1},"newtons-mirror":{"score":10.1,"rawScore":10.1,"count":1,"countedResults":1,"prompts":1},"ground-truth":{"score":20.2,"rawScore":20.2,"count":1,"countedResults":1,"prompts":1},"raccoon":{"score":36.7,"rawScore":36.7,"count":1,"countedResults":1,"prompts":1},"pelican":{"score":38,"rawScore":38,"count":1,"countedResults":1,"prompts":1},"foresight-self-improvement":{"score":3.2,"rawScore":3.2,"count":1,"countedResults":1,"prompts":1},"latch-five-levels":{"score":3.6,"rawScore":3.6,"count":1,"countedResults":1,"prompts":1}},"capabilities":{"general":{"score":11.9,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":11.3,"upper":12.4},"availableFamilies":8,"expectedFamilies":8,"missingFamilies":[],"partial":false},"visual":{"score":32.2,"uncertainty":{"scope":"score-resolution-and-missing-scenes","lower":30.1,"upper":34.3},"availableFamilies":6,"expectedFamilies":6,"missingFamilies":[],"partial":false}}}},"reasoning":{"singleStep":{"score":27},"multiStep":{"score":47}}}],"results":[{"id":"space-bunny-alpha-svg-selfie-2a4532da99dfad0f","modelId":"space-bunny-alpha","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An astronaut rabbit reaches for a carrot above a small moon while a younger companion peeks into view. The scene tests expressive characters, spatial relationships and material detail.","promptHash":"2a4532da99dfad0f","date":"","edition":{"id":"before-20260929","label":"Before 29 Sep"},"media":{"type":"image","src":"/benchmarks/media/864f2a16aa6a6438.png","alt":"Space Bunny Alpha: Selfie SVG benchmark"},"quality":{"score":80.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"An appealing storybook illustration with expressive eyes, coherent limbs, pearly shading and carefully layered accessories. The requested moon, carrot, smaller rabbit and floating rocks are clear. Limits include a wider-than-tall canvas, smooth rather than plush fur, weak push-off contact, indistinct raised fingers and an angular turn in the orbital ring. Surfaces remain visibly stylized.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":8},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An astronaut rabbit reaches for a carrot above a small moon while a younger companion peeks into view. The scene tests expressive characters, spatial relationships and material detail.","promptSummary":"An astronaut rabbit reaches for a carrot above a small moon while a younger companion peeks into view. The scene tests expressive characters, spatial relationships and material detail.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-mona-lisa-012e622dada8da31","modelId":"space-bunny-alpha","groupId":"mona-lisa","kind":"svg","title":"mona-lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"012e622dada8da31","date":"","edition":{"id":"before-20260929","label":"Before 29 Sep"},"media":{"type":"image","src":"/benchmarks/media/7f15491eeff9d5c4.png","alt":"Space Bunny Alpha: Mona Lisa SVG benchmark"},"quality":{"score":48.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered portrait, dark clothing, crossed hands and winding landscape clearly recall the Mona Lisa. Replication is limited by the oversized, rounded face, frontal pose and simplified expression. Wrists connect to sleeves, but crowded fingers and flattened palms weaken the anatomy. Soft facial shading adds volume; angular scenery and repetitive garment lines limit painterly depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-pelican-dc9481a53c0193ac","modelId":"space-bunny-alpha","groupId":"pelican","kind":"svg","title":"pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"dc9481a53c0193ac","date":"","edition":{"id":"before-20260929","label":"Before 29 Sep"},"media":{"type":"image","src":"/benchmarks/media/3f27675de08442f3.png","alt":"Space Bunny Alpha: Pelican SVG benchmark"},"quality":{"score":79.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"The recognizable pelican, expressive bill and balanced lakeside composition make an appealing illustration. Connected legs and a wing reaching the handlebar establish the riding action. The steering stem and fork change direction awkwardly, while crowded foot/crank overlaps reduce mechanical clarity. Feather contours and spokes reward inspection; shallow shading and limited directional shadows constrain depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-raccoon-6ee169059ae55c1a","modelId":"space-bunny-alpha","groupId":"raccoon","kind":"svg","title":"raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"6ee169059ae55c1a","date":"","edition":{"id":"before-20260929","label":"Before 29 Sep"},"media":{"type":"image","src":"/benchmarks/media/a13c80b6057c9078.png","alt":"Space Bunny Alpha: Raccoon SVG benchmark"},"quality":{"score":78.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"The raccoon clearly rides on the dog's back, and the dog's paws reach a connected steering assembly. Expressive faces, controlled outlines and layered mountains make an appealing illustration. The seated anatomy and hull geometry are simplified, while the sunset has little corresponding reflection or directional shading. Curved spray conveys motion, although several splash ribbons look decorative.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-beach-ecf1075b5d6428a9","modelId":"space-bunny-alpha","groupId":"beach","kind":"svg","title":"beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"ecf1075b5d6428a9","date":"","edition":{"id":"before-20260929","label":"Before 29 Sep"},"media":{"type":"image","src":"/benchmarks/media/bc67f9a7c79036c5.png","alt":"Space Bunny Alpha: Beach SVG benchmark"},"quality":{"score":75.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"An appealing sunset beach illustration with recognizable animals, birds, playful accessories and a clear sun reflection. Layered waves and warm-to-cool colors organize the scene effectively. Animal construction is simplified, the turtle's flippers are awkward, and foreground objects receive limited sunset illumination. Clean vector shapes provide polish, but volume, contact and material variation remain modest.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"pixel-canary-svg-selfie-ad78b3d2cc9de9e7","modelId":"pixel-canary","groupId":"selfie","kind":"svg","title":"Canary selfie","prompt":"A rounded bird built from small cubes explores a miniature machine. The scene tests block construction, expressive proportions, contact and soft lighting.","promptHash":"ad78b3d2cc9de9e7","date":"2026-09-26","generatedAt":"2026-09-26","media":{"type":"image","src":"/benchmarks/media/d6827ba6d352bfec.png","alt":"Pixel Canary’s SVG self-portrait: a round golden voxel canary placing a cyan block into a miniature control panel."},"firstGeneratedAt":"2026-09-26","quality":{"score":74.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The cheerful yellow bird, cream eyes, orange features and miniature teal console closely match the brief. Clean block edges, restrained colors and soft shadows create an appealing composition. The body reads more like a gridded disc than a fully cubic sphere, and the outlined left wing has limited volume. The right wing reaches the cyan block, though its grip is ambiguous.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A rounded bird built from small cubes explores a miniature machine. The scene tests block construction, expressive proportions, contact and soft lighting.","promptSummary":"A rounded bird built from small cubes explores a miniature machine. The scene tests block construction, expressive proportions, contact and soft lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"pixel-canary-svg-mona-lisa-63163c6b35a3998f","modelId":"pixel-canary","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-09-26","generatedAt":"2026-09-26","media":{"type":"image","src":"/benchmarks/media/16c62f660e51d288.png","alt":"Pixel Canary’s SVG interpretation of the Mona Lisa, with soft painted shading and the complete ornate gold frame."},"firstGeneratedAt":"2026-09-26","quality":{"score":45.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The framed portrait recalls Mona Lisa through dark hair, folded arms and a hazy landscape. Replication is limited by wide, staring eyes, an elongated neck and a downturned mouth. The overlapping hands lack convincing finger and wrist construction. Soft tonal shading suggests volume, but pervasive blur obscures clothing, landscape and anatomical detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":3.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":-0.06856536143963214,"y":0.0057915290759217326,"scale":1.2336780599567423},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"pixel-canary-svg-pelican-f998b992640eff5e","modelId":"pixel-canary","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-09-26","generatedAt":"2026-09-26","media":{"type":"image","src":"/benchmarks/media/29e0c94a1089960a.svg","alt":"Pixel Canary: Pelican SVG benchmark","still":"/benchmarks/media/e1ffaf6df28fef77.svg"},"firstGeneratedAt":"2026-09-26","quality":{"score":76.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"An appealing cartoon with a distinctive pelican, expressive eye, warm bill and carefully drawn bicycle wheels. The wing reaches the handlebar, but the feet do not clearly engage the pedals, weakening the riding mechanics. Layered hills create depth; limited ground shadows weaken contact. Clean curves and purposeful feather details give the illustration polish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"pixel-canary-svg-raccoon-923e67d1bd1a4fda","modelId":"pixel-canary","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"923e67d1bd1a4fda","date":"2026-09-26","generatedAt":"2026-09-26","media":{"type":"image","src":"/benchmarks/media/776595725a8a3a21.png","alt":"Pixel Canary’s SVG of a raccoon holding a drink on a goggle-wearing dog’s back as the dog steers a jet ski across the sea."},"firstGeneratedAt":"2026-09-26","quality":{"score":77.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon clearly rides the dog's back, and the dog's paws reach the jet ski controls. Expressive faces, a sweeping tail and a strong sunset composition give the illustration character. Limbs remain connected, but the dog's foreleg joints and vehicle seating are loosely constructed. Angular splashes convey speed more successfully than convincing water displacement; lighting and reflections remain decorative.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":2},{"id":"pixel-canary-svg-beach-48eb1c58439ea576","modelId":"pixel-canary","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-09-26","generatedAt":"2026-09-26","media":{"type":"image","src":"/benchmarks/media/4e74ae8a736132c5.png","alt":"Pixel Canary’s SVG beach sunset with dolphins, birds, a kite, beachgoers, a crab, a turtle, a sandcastle and reflections in the water."},"firstGeneratedAt":"2026-09-26","quality":{"score":61.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cohesive sunset beach illustration with lively wildlife, playful accents, and an appealing warm-to-purple palette. Palm framing and the aligned sun reflection organize the scene, while connected silhouettes keep the activities readable. Schematic joints, ambiguous dolphin junctions, loosely matched reflections, and repetitive surface marks limit construction, lighting coherence, and material depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"ox-alpha-svg-selfie-886d15bfe4c213b1","modelId":"ox-alpha","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A rounded jade bull raising a hoof tests anatomy, balance and the appearance of polished stone and metal.","promptHash":"886d15bfe4c213b1","date":"2026-09-27","generatedAt":"2026-09-27T02:13:40.024Z","media":{"type":"image","src":"/benchmarks/media/c6567cdb9ff92192.png","alt":"Ox Alpha: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-27T01:44:23.101Z","quality":{"score":76.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The expressive emerald face, ivory horns, gold accents and fine veining closely match the brief. Rounded volumes and layered shading create an appealing mascot. Rear hoof placement is cramped, lighting is somewhat inconsistent, and conspicuous white patches behind the muzzle interrupt the finish. Surfaces remain visibly stylized rather than photographic.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A rounded jade bull raising a hoof tests anatomy, balance and the appearance of polished stone and metal.","promptSummary":"A rounded jade bull raising a hoof tests anatomy, balance and the appearance of polished stone and metal.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"ox-alpha-svg-mona-lisa-63163c6b35a3998f","modelId":"ox-alpha","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-09-27","generatedAt":"2026-09-27T02:33:46.492Z","media":{"type":"image","src":"/benchmarks/media/8a36c1e023b9832b.png","alt":"Ox Alpha: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-27T02:33:46.492Z","quality":{"score":49.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The crossed arms, green dress and hazy landscape clearly evoke Mona Lisa. Replication is limited by the frontal pose, oversized eyes, elongated face and pronounced lips. Overlapping finger strips lack convincing knuckles and palm volume. The balanced framing and soft shading provide cohesion, but simplified anatomy, repetitive folds and blurred scenery keep the portrait basic.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.03583506510066377,"y":0.07127026198834792,"scale":1.1972173631218097},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"ox-alpha-svg-pelican-1070dc01fd3bbee4","modelId":"ox-alpha","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"1070dc01fd3bbee4","date":"2026-09-27","generatedAt":"2026-09-27T02:48:25.188Z","media":{"type":"image","src":"/benchmarks/media/baedf5d9bb6a37b3.png","alt":"Ox Alpha: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-27T02:48:25.188Z","quality":{"score":75.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The pelican, bicycle and seaside setting form a clear, appealing illustration. Circular wheels, readable frame tubes and clean silhouettes support the action. The legs connect to the body, but the far foot's pedal contact and wing-to-handlebar junction are cramped. Simple shading, uniform wave marks and limited surface texture keep the rendering relatively flat.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"ox-alpha-svg-beach-48eb1c58439ea576","modelId":"ox-alpha","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-09-27","generatedAt":"2026-09-27T02:48:25.374Z","media":{"type":"image","src":"/benchmarks/media/480040d5de9d4533.png","alt":"Ox Alpha: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-27T02:48:25.374Z","quality":{"score":77.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Framing palms, a vivid sunset palette and balanced kite and umbrella create an appealing beach illustration. Layered water and shoreline give the scene depth, while aligned reflections reinforce the sunset. Small accessories add character and reward inspection. Figures and animals remain schematic, some overlaps are difficult to read, and simplified shadows and flat surfaces limit physical richness.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"owl-alpha-svg-selfie-e311e3b689f8e6f2","modelId":"owl-alpha","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A cloaked owl with luminous eyes tests expression, fabric folds and the separation of dark forms.","promptHash":"e311e3b689f8e6f2","date":"2026-09-28","generatedAt":"2026-09-28T02:04:49.187Z","media":{"type":"image","src":"/benchmarks/media/ad3b9ff625b0748a.png","alt":"Owl Alpha: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-28","quality":{"score":61.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"Amber slit eyes, a pointed beak, cream feathers and gold-trimmed black clothing establish the secretive owl rogue. The head appears frontal rather than lowered, with much of the face exposed. The centered silhouette is effective, but stiff cloak geometry, shallow facial modeling and extremely dark clothing limit depth and finish. Sparse gold markings suggest the requested circuitry only loosely.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A cloaked owl with luminous eyes tests expression, fabric folds and the separation of dark forms.","promptSummary":"A cloaked owl with luminous eyes tests expression, fabric folds and the separation of dark forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"owl-alpha-svg-mona-lisa-63163c6b35a3998f","modelId":"owl-alpha","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-09-28","generatedAt":"2026-09-28T02:05:06.913Z","media":{"type":"image","src":"/benchmarks/media/4246b768abbdd251.png","alt":"Owl Alpha: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-28","quality":{"score":4.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered portrait, slight smile, overlapping hands and distant landscape suggest Mona Lisa, but replication is weak. The face has little volume, the shoulders and torso form a narrow wedge, and the hands are fingerless ovals with unresolved wrists. Broad gray background bands provide limited depth. A bright head rim and pale hands contrast with an almost featureless dark figure.","categories":[{"id":"fidelity","label":"Task & likeness","score":0.5},{"id":"structure","label":"Anatomy & construction","score":0.5},{"id":"space","label":"Space & placement","score":0.5},{"id":"light","label":"Light & reflections","score":0.5},{"id":"detail","label":"Purposeful detail","score":0},{"id":"artistry","label":"Artistry & character","score":0.5},{"id":"effects","label":"Materials & effects","score":0},{"id":"integrity","label":"Visual integrity","score":0.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":-0.011462597318995389,"y":0.20904792894609864,"scale":1.0304545339535158},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"owl-alpha-svg-pelican-1070dc01fd3bbee4","modelId":"owl-alpha","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"1070dc01fd3bbee4","date":"2026-09-28","generatedAt":"2026-09-28T02:37:39.545Z","media":{"type":"image","src":"/benchmarks/media/fab029fc937f0e9b.png","alt":"Owl Alpha: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-28","quality":{"score":79,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"A clearly recognizable pelican cycles through a well-composed coastal sunset. Warm colors, flowing scarf, feather contours and bicycle details give the illustration character. The legs connect to the body, but the farther foot's pedal contact and the wing's handlebar grip remain ambiguous. Clean vector shapes provide polish, with limited material texture and directional shading.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"owl-alpha-svg-raccoon-7e17566f15394e20","modelId":"owl-alpha","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-09-28","generatedAt":"2026-09-28T02:30:40.495Z","media":{"type":"image","src":"/benchmarks/media/68631133c65eed0e.png","alt":"Owl Alpha: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-28","quality":{"score":43.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"The raccoon is recognizable and expressive, with a clear mask and striped tail. The dog's anatomy is poorly resolved, and the flattened craft lacks convincing steering and seating. Feet extend below the hull, weakening the sense of support. An appealing palette, clean shapes and decorative water effects give the illustration some polish.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"owl-alpha-svg-beach-c148a8dd4078bd15","modelId":"owl-alpha","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"c148a8dd4078bd15","date":"2026-09-28","generatedAt":"2026-09-28T02:05:08.616Z","media":{"type":"image","src":"/benchmarks/media/c8c6efed87832aa3.png","alt":"Owl Alpha: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-28","quality":{"score":54.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"The sunset, palms, birds and beach animals create a cheerful, readable scene. Clean silhouettes and a vivid sky establish the composition, but construction remains schematic. The right palm appears rooted in open water, shadows have limited connection to the sunset, and the sun’s reflection consists of rigid bands. Small creatures and beach equipment are recognizable but minimally developed.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":3},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"union-alpha-svg-pelican-b60df10b578f3008","modelId":"union-alpha","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b60df10b578f3008","date":"","media":{"type":"image","src":"/benchmarks/media/d502c503b55f80bd.png","alt":"A pelican riding a teal bicycle beneath a sunny blue sky."},"quality":{"score":67,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The pelican is immediately recognizable, with an expressive face, convincing bill silhouette and appealing pastel composition. Both legs connect to feet, and the wing reaches the handlebar. Bicycle construction is less resolved: the far pedal lacks a clear crank connection and no chain is visible. Clean contours and orderly spokes support the illustration, while flat lighting and simplified surfaces limit depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"union-alpha-svg-raccoon-f6c8698d226f646f","modelId":"union-alpha","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"f6c8698d226f646f","date":"","media":{"type":"image","src":"/benchmarks/media/814f7e54d6c415cc.png","alt":"A raccoon rides on a dog as the dog steers a red jet ski."},"quality":{"score":70.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Expressive animals, a readable riding arrangement and sweeping spray create an appealing cartoon. The raccoon sits convincingly on the dog's back, but the dog's paw does not clearly grip the handlebar, whose connection to the hull is unresolved. Flat lighting provides little volume or reflection. A dark rectangular strip interrupts the upper-right background.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"union-alpha-svg-mona-lisa-012e622dada8da31","modelId":"union-alpha","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"012e622dada8da31","date":"","media":{"type":"image","src":"/benchmarks/media/4e06880a401d846b.png","alt":"An SVG interpretation of the Mona Lisa, with folded hands and a landscape background."},"quality":{"score":59.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered portrait, folded hands, dark clothing and hazy landscape clearly evoke the Mona Lisa. Muted colors and soft facial shading create a coherent illustration, but replication remains schematic: the face is broad, eyes outlined, smile simplified and overlapping fingers poorly resolved. Angular landscape shapes and flat fabric limit depth and painterly richness.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.037666500078144406,"y":0.11149099640134821,"scale":1.3297620281214726},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"union-alpha-svg-beach-ecf1075b5d6428a9","modelId":"union-alpha","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"ecf1075b5d6428a9","date":"","media":{"type":"image","src":"/benchmarks/media/dac1b0b240bfa76c.png","alt":"A beach at sunset with a palm tree, turtle, crab, birds, and a leaping dolphin."},"quality":{"score":71.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cohesive sunset palette, sweeping shoreline and playful animals make an appealing beach illustration. The sun’s reflection establishes a clear focal axis, balanced by the palm and umbrella. Clean silhouettes and small decorative features reward inspection, though simplified anatomy, repetitive reflection bars and generic oval shadows limit volume and naturalism.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-selfie-2a4532da99dfad0f-from-20260929","modelId":"space-bunny-alpha","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An astronaut rabbit reaches for a carrot above a small moon while a younger companion peeks into view. The scene tests expressive characters, spatial relationships and material detail.","promptHash":"2a4532da99dfad0f","date":"2026-09-29","generatedAt":"2026-09-29T19:05:43.624Z","edition":{"id":"from-20260929","label":"29 Sep onward"},"media":{"type":"image","src":"/benchmarks/media/e2b5ef9567d098aa.png","alt":"Space Bunny Alpha: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-29T19:05:43.624Z","quality":{"score":78.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The expressive rabbits, lavender moon and diagonal ring create an appealing storybook composition with nearly all requested features. Glossy eyes and pearly highlights work well. The helmet leaves the towering ears unenclosed, the reaching arm has little joint definition, and the spread boots suggest hovering more clearly than pushing off. Repeated scalloped fur marks and broad gradients limit material depth; purple edge halos soften otherwise controlled silhouettes.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An astronaut rabbit reaches for a carrot above a small moon while a younger companion peeks into view. The scene tests expressive characters, spatial relationships and material detail.","promptSummary":"An astronaut rabbit reaches for a carrot above a small moon while a younger companion peeks into view. The scene tests expressive characters, spatial relationships and material detail.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-mona-lisa-012e622dada8da31-from-20260929","modelId":"space-bunny-alpha","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"012e622dada8da31","date":"2026-09-29","generatedAt":"2026-09-29T19:13:17.233Z","edition":{"id":"from-20260929","label":"29 Sep onward"},"media":{"type":"image","src":"/benchmarks/media/4812db5e8e91d920.png","alt":"Space Bunny Alpha: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T19:13:17.233Z","quality":{"score":49.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The dark clothing, folded hands and muted landscape make the reference recognizable. Replication is limited by the broad frontal face, oversized head, explicit smile and simplified landscape. Arms connect coherently, but overlapping hands lack convincing finger construction. Soft gradients suggest volume and atmospheric depth; clothing folds and facial features remain schematic.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.01808955897874828,"y":0.007952513724921718,"scale":1.4333294145603397},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-pelican-dc9481a53c0193ac-from-20260929","modelId":"space-bunny-alpha","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"dc9481a53c0193ac","date":"2026-09-29","generatedAt":"2026-09-29T19:13:18.847Z","edition":{"id":"from-20260929","label":"29 Sep onward"},"media":{"type":"image","src":"/benchmarks/media/5d0c5ffbce3f9b2b.png","alt":"Space Bunny Alpha: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T19:13:18.847Z","quality":{"score":68,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The large pouched bill clearly identifies the pelican, and the balanced bicycle silhouette makes the action readable. Attached legs reach the crank area, but overlapping feet and crank lines leave pedal contact ambiguous. Consistent wheel scale, clean contours and a restrained coastal palette support an appealing composition. Lighting and surface treatment remain simple and mostly flat.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-beach-ecf1075b5d6428a9-from-20260929","modelId":"space-bunny-alpha","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"ecf1075b5d6428a9","date":"2026-09-29","generatedAt":"2026-09-29T19:13:20.727Z","edition":{"id":"from-20260929","label":"29 Sep onward"},"media":{"type":"image","src":"/benchmarks/media/abfa0e6d05ae8c9e.png","alt":"Space Bunny Alpha: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T19:13:20.727Z","quality":{"score":67.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cheerful, coherent beach illustration with a strong sunset focal point, varied animals and readable small accessories. Connected flamingo legs and grounded foreground objects support the scene. Construction remains schematic, especially the airborne marine animal; its jump lacks clear water displacement. The sun reflection is effective but geometrically rigid, while shadows and material textures provide limited depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-svg-mona-lisa-e7b85d8cd393c26f","modelId":"deepseek/deepseek-v4.1-flash","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-09-29","generatedAt":"2026-09-29T16:51:16.628Z","media":{"type":"image","src":"/benchmarks/media/7f0e6c31ef7cb943.png","alt":"DeepSeek-V4.1-Flash: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T16:51:16.628Z","quality":{"score":37.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered portrait, dark hair, crossed hands and muted landscape make the reference recognizable. Replication is limited by the frontal pose, staring eyes and weakly defined smile. Hands are overlapping oval masses with almost no finger articulation, while wrists disappear into sleeves. Soft gradients suggest volume, but broad, indistinct surfaces and faint stray lines leave the drawing basic and uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":3.5},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":3},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04448390597200633,"y":0.19265635505679965,"scale":1.3909681284637785},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-svg-pelican-b618cda0f539e818","modelId":"deepseek/deepseek-v4.1-flash","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T16:51:16.628Z","media":{"type":"image","src":"/benchmarks/media/6d4a621a488b7539.png","alt":"DeepSeek-V4.1-Flash: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T16:51:16.628Z","quality":{"score":75,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"A recognizable pelican rides a brightly colored bicycle against a striking sunset. The sweeping bill, trailing scarf and circular wheels create a controlled composition. Both legs connect, but pedal linkage, body support and the wing's handlebar grip remain loosely resolved. Clean vector shapes and feather details work well; simplified shading limits volume and integration with the sunset light.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-svg-raccoon-f360071d514152be","modelId":"deepseek/deepseek-v4.1-flash","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"f360071d514152be","date":"2026-09-29","generatedAt":"2026-09-29T16:54:55.087Z","media":{"type":"image","src":"/benchmarks/media/01c2c86112476d85.png","alt":"DeepSeek-V4.1-Flash: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T16:54:55.087Z","quality":{"score":43.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The sunset palette and silhouetted grouping establish an appealing scene. A masked rider sits above a pointed-eared dog, but merged body shapes leave limbs and support unresolved. The boatlike craft lacks readable steering controls, so the requested action is unclear. Sparse facial marks and glowing spray dots offer limited detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-svg-beach-d030aa6f45fe19e9","modelId":"deepseek/deepseek-v4.1-flash","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T16:51:16.628Z","media":{"type":"image","src":"/benchmarks/media/e1665469bde4b0bb.png","alt":"DeepSeek-V4.1-Flash: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T16:51:16.628Z","quality":{"score":59.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The glowing sunset, silhouetted palm and playful foreground create a coherent, appealing beach illustration. Birds, crab, boat and reflections fulfill the brief. The airborne animals have weakly defined anatomy and little convincing splash interaction. Broad, blurred reflections and largely flat surfaces limit depth and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-pro-0813-svg-mona-lisa-e7b85d8cd393c26f","modelId":"deepseek/deepseek-v4-pro-0813","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-09-29","generatedAt":"2026-09-29T17:22:13.467Z","media":{"type":"image","src":"/benchmarks/media/a0a29e2ffa8f7fbc.png","alt":"DeepSeek-V4-Pro-0813: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T17:22:13.467Z","quality":{"score":16.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"Dark hair, a restrained smile and a hazy landscape make the reference recognizable. Severe construction problems dominate: oversized looping sleeves surround a background-colored opening, and the crossed hands lack coherent wrist and forearm connections. Rigid symmetry and elongated facial proportions weaken expression. Soft skin gradients and clean contours provide limited finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":2},{"id":"structure","label":"Anatomy & construction","score":0.5},{"id":"space","label":"Space & placement","score":2},{"id":"light","label":"Light & reflections","score":2},{"id":"detail","label":"Purposeful detail","score":2},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":1.5},{"id":"integrity","label":"Visual integrity","score":1}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04294258701272386,"y":0.09941053646547367,"scale":1.4549914146181997},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-pro-0813-svg-pelican-b618cda0f539e818","modelId":"deepseek/deepseek-v4-pro-0813","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T17:22:14.106Z","media":{"type":"image","src":"/benchmarks/media/c297396d130ced18.png","alt":"DeepSeek-V4-Pro-0813: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T17:22:14.106Z","quality":{"score":62.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"A clearly recognizable pelican cycles through a balanced, expressive illustration. The sweeping bill, shaded plumage and carefully drawn wheels give it character. Mechanical construction is weaker: the chain loop sits below the crank, and foot-to-pedal and wing-to-handlebar contact are ambiguous. Smooth gradients provide volume, but the scene remains visibly stylized.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-pro-0813-svg-raccoon-b4cedf42a4a26a4b","modelId":"deepseek/deepseek-v4-pro-0813","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T17:22:13.481Z","media":{"type":"image","src":"/benchmarks/media/b2127cf4cff85cc6.png","alt":"DeepSeek-V4-Pro-0813: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T17:22:13.481Z","quality":{"score":54,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The masked raccoon riding the steering dog is readily recognizable, with expressive faces and energetic diagonal movement. The oversized tail dominates the composition, limb connections are difficult to trace, and the jet ski has little structural depth. Blurred spray and a heavy oval shadow weaken water contact, while the glowing sun produces little corresponding illumination.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-pro-0813-svg-beach-d030aa6f45fe19e9","modelId":"deepseek/deepseek-v4-pro-0813","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T17:22:17.363Z","media":{"type":"image","src":"/benchmarks/media/4b6b34dc99cae651.png","alt":"DeepSeek-V4-Pro-0813: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T17:22:17.363Z","quality":{"score":48.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"The glowing sunset, framing palms and layered shoreline create a coherent beach scene with animals, birds and playful accessories. The airborne blue animal has unclear anatomy, while flat foliage, weak ground contact and broad oval reflections limit depth and finish. The image is appealing but visibly simplified.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":3.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-flash-svg-mona-lisa-e7b85d8cd393c26f","modelId":"qwen/qwen3.8-flash","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-09-29","generatedAt":"2026-09-29T18:13:18.970Z","media":{"type":"image","src":"/benchmarks/media/e153707508afe6c4.png","alt":"Qwen3.8-Flash: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T18:13:18.970Z","quality":{"score":51.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The dark dress, crossed hands and hazy landscape evoke the Mona Lisa, but replication is limited by a generic frontal face, coarse smile and poorly articulated fingers. The gold frame has clearer definition than the portrait. Heavy mottling obscures features, while soft shading provides some volume without convincing anatomical detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.03887205872897426,"y":-0.013937451875664397,"scale":1.1972173631218093},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-flash-svg-pelican-b618cda0f539e818","modelId":"qwen/qwen3.8-flash","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T17:46:16.674Z","media":{"type":"image","src":"/benchmarks/media/c42adec29a147f11.png","alt":"Qwen3.8-Flash: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T17:46:16.674Z","quality":{"score":77.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The pelican's expressive face, oversized bill and flowing scarf create an appealing focal point. The bicycle has recognizable frame, chain and wheel construction, but the far foot lacks a clear pedal connection and the wing's steering contact is ambiguous. Clean contours and gentle gradients support the illustration; simplified volume and weak directional shadows limit depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-flash-svg-raccoon-b4cedf42a4a26a4b","modelId":"qwen/qwen3.8-flash","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T18:13:19.760Z","media":{"type":"image","src":"/benchmarks/media/5ba23f0c66e43ca0.png","alt":"Qwen3.8-Flash: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T18:13:19.760Z","quality":{"score":62.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon clearly rides the dog's back, with expressive features and appealing blue-orange color contrast. The dog's paws do not clearly operate connected controls, and the shallow vehicle lacks convincing jet ski construction. Clean outlines support the cartoon style, but generic shading and blurred water streaks limit depth and motion.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-flash-svg-beach-d030aa6f45fe19e9","modelId":"qwen/qwen3.8-flash","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T17:46:16.674Z","media":{"type":"image","src":"/benchmarks/media/58f05d3c275ab256.png","alt":"Qwen3.8-Flash: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T17:46:16.674Z","quality":{"score":77.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Sweeping palms, coordinated sunset colors and a central reflection create an expressive beach composition with clear depth. Birds, animals and varied activities thoroughly realize the brief. Small figures and animal anatomy remain rudimentary, water movement is schematic, and contact shadows are weak. Scattered foreground details and the clipped orange disk limit the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-1-sol-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-6.1-sol","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T18:28:54.385Z","media":{"type":"image","src":"/benchmarks/media/7d08d3b889d5323d.png","alt":"GPT-6.1 Sol: Mona Lisa SVG benchmark","thumbnail":"/assets/homepage/thumbnails/b9fb9ee8e81a3690a8ee43cb.png"},"firstGeneratedAt":"2026-09-29T18:28:54.385Z","quality":{"score":61.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The recognizable pose, dark clothing and misty landscape form a coherent portrait. Soft facial shading creates volume, but the broad face, outlined eyes and simplified smile limit likeness. Overlapping hands have flattened fingers and ambiguous separation. Angular landscape shapes and sparse fabric detail keep the execution competent rather than masterful.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":-0.0016665390590095994,"y":0.10923787239555023,"scale":1.3909681284637796},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-1-sol-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-6.1-sol","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T18:28:54.385Z","media":{"type":"image","src":"/benchmarks/media/06ca2b4636365059.png","alt":"GPT-6.1 Sol: Pelican SVG benchmark","thumbnail":"/assets/homepage/thumbnails/bcc86a42bda30eb6fadfc40a.png"},"firstGeneratedAt":"2026-09-29T18:28:54.385Z","quality":{"score":79,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"A clearly recognizable pelican rides a well-composed bicycle through a warm coastal scene. The expressive bill, restrained palette and carefully drawn wheels give the illustration character. Leg-to-pedal connections are less convincing, especially the obscured far foot. Lighting provides gentle volume but little directional shadow. The small subtitle is clipped at the lower border.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-1-sol-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-6.1-sol","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T18:28:54.385Z","media":{"type":"image","src":"/benchmarks/media/51bbdfc6e7c92033.png","alt":"GPT-6.1 Sol: Raccoon SVG benchmark","thumbnail":"/assets/homepage/thumbnails/0f86467d6db6f7fcc13b4dbf.png"},"firstGeneratedAt":"2026-09-29T18:28:54.385Z","quality":{"score":77.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"The masked raccoon rides the dog's back, and the dog's forepaw meets a handlebar connected to the deck. Expressive faces, a striped tail and sweeping wake create an appealing composition. Overlapping bodies compress the seating and rear-leg anatomy; hull depth and sunset illumination remain schematic.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-1-sol-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-6.1-sol","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T18:28:54.386Z","media":{"type":"image","src":"/benchmarks/media/53c33e28af2c873f.png","alt":"GPT-6.1 Sol: Beach SVG benchmark","thumbnail":"/assets/homepage/thumbnails/a89b2ae0d51596b366cc0e1c.png"},"firstGeneratedAt":"2026-09-29T18:28:54.386Z","quality":{"score":76.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cohesive sunset palette, sweeping shoreline and framing palm create an appealing beach illustration. Animals, birds, play and a prominent sun reflection fulfill the brief. Connected but schematic figures and simplified dolphin anatomy limit construction; stepped reflections and mostly flat surfaces limit lighting and material depth. Small beach accessories remain readable and edges are clean.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-raccoon-534827245194ac5e-from-20260929","modelId":"space-bunny-alpha","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"534827245194ac5e","date":"2026-09-29","generatedAt":"2026-09-29T19:53:54.577Z","edition":{"id":"from-20260929","label":"29 Sep onward"},"media":{"type":"image","src":"/benchmarks/media/9b6b3bd6e22ecb8c.png","alt":"Space Bunny Alpha: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T19:41:54.241Z","quality":{"score":73.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon clearly rides on the dog's back, and the dog's extended paw reaches the steering bar. Expressive faces, balanced silhouettes and crisp outlines make an appealing illustration. The crowded steering assembly and flattened cockpit geometry weaken construction; lighting and materials remain simplified despite effective splashes.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-sol-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-5.6-sol","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.325Z","media":{"type":"image","src":"/benchmarks/media/40f78ddb6556530d.png","alt":"GPT-5.6 Sol: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.325Z","quality":{"score":48,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The crossed hands, dark dress and winding landscape establish the Mona Lisa composition. Replication is limited by enlarged eyes, an elongated face and a more frontal pose. Broad gradients suggest volume, but outlined features and flattened, overlapping fingers weaken the portrait. The frame and restrained palette unify the image; visible grain adds texture without convincing painterly transitions.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.03663857076360762,"y":0.09641878349084242,"scale":1.6404982390570424},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-sol-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-5.6-sol","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.325Z","media":{"type":"image","src":"/benchmarks/media/559fbb5436ca212c.png","alt":"GPT-5.6 Sol: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.325Z","quality":{"score":67.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The expressive pelican and balanced coastal composition remain appealing, but a sky-colored slit between the neck contours above the wing leaves the central anatomy incoherent. This visible construction and layering defect prevents a clean-integrity assessment. Clustered feet lack clear pedal contacts and the chain routing is schematic; lighting and water reflections remain graphic.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-sol-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-5.6-sol","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.325Z","media":{"type":"image","src":"/benchmarks/media/499c0bd2f4d480b7.png","alt":"GPT-5.6 Sol: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.325Z","quality":{"score":62.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Expressive faces, bold silhouettes and layered waves make an appealing cartoon. The raccoon sits convincingly behind the dog's shoulders, and the dog's forward paw reaches the steering grip. Steering construction and seated support remain somewhat ambiguous. The glowing sunset lacks a corresponding warm reflection or consistent illumination on the riders.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-sol-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-5.6-sol","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.325Z","media":{"type":"image","src":"/benchmarks/media/ba46d329d0c80fe9.png","alt":"GPT-5.6 Sol: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.325Z","quality":{"score":68,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-12","rationale":"The purple-to-gold sunset, framing palms and varied beach activities create an appealing, coherent illustration. The sun’s reflection establishes a clear focal path. Figures and animals are recognizable but geometrically simplified; airborne dolphins lack convincing splashes, and the boat reflection is schematic. Clean silhouettes and playful details succeed more strongly than anatomical depth or material realism.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-terra-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-5.6-terra","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.489Z","media":{"type":"image","src":"/benchmarks/media/95921f82ea30748a.png","alt":"GPT-5.6 Terra: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.489Z","quality":{"score":41.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The dark clothing, central portrait and restrained smile evoke the Mona Lisa, but replication is limited by the broad face, elongated neck, nearly frontal pose and indistinct landscape. Soft gradients give the face some volume. Crossed arms become rounded masses without readable fingers or wrists, and the expression lacks the original's subtlety. Clean framing and curves support an otherwise basic, uneven portrait.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05634624923335067,"y":0.3503029101291843,"scale":2.247907986676468},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-terra-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-5.6-terra","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.489Z","media":{"type":"image","src":"/benchmarks/media/556d9325c3053db9.png","alt":"GPT-5.6 Terra: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.489Z","quality":{"score":65,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The oversized bill, expressive eye and balanced twilight palette make the cyclist immediately readable. Clean contours and wheel details support the illustration, but the rabbit-like head tufts weaken pelican likeness. The wing does not clearly grip the handlebar, pedal contact is uneven, and the drivetrain is incomplete. Lighting remains mostly decorative.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-terra-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-5.6-terra","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.489Z","media":{"type":"image","src":"/benchmarks/media/0d34d6329c7e9a61.png","alt":"GPT-5.6 Terra: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.489Z","quality":{"score":63.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Expressive faces, a recognizable raccoon perched on the dog, and a sweeping wake make an appealing cartoon. The principal weakness is the steering interaction: the dog's paw sits beside the control assembly without a convincing handlebar grip. Bold outlines and coherent layering provide polish, while simplified shading, compressed seating geometry and angular splashes limit depth and physical credibility.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-terra-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-5.6-terra","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:39:48.489Z","media":{"type":"image","src":"/benchmarks/media/cfcafe026b5bab63.png","alt":"GPT-5.6 Terra: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:39:48.489Z","quality":{"score":56.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Warm sunset gradients, framing palms and colorful beach activities create an appealing scene. Animals, birds and a sun reflection fulfill the brief. The dog and turtle remain schematic, foreground depth is shallow, and the reflection forms rigid bands. Clean silhouettes and readable accessories support a competent illustration, but limited anatomy, material variation and lighting keep it below exceptional.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-sol-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-6-sol","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T22:40:29.443Z","media":{"type":"image","src":"/benchmarks/media/186d400bbafeb8e7.png","alt":"GPT-6 Sol: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:40:29.443Z","quality":{"score":48.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The dark clothing, crossed arms, restrained smile and distant landscape establish the Mona Lisa reference. Replication is limited by the oversized, nearly frontal head, schematic eyes and fingerless hand shapes. Soft shading suggests facial volume, but muddy transitions and angular scenery weaken depth and painterly finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.08526940838867066,"y":0.33139019706499484,"scale":1.9347923344020284},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-sol-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-6-sol","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:42:45.496Z","media":{"type":"image","src":"/benchmarks/media/f6ea8e1340bd341f.png","alt":"GPT-6 Sol: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:42:45.496Z","quality":{"score":80.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"The pelican reads clearly, with an expressive face, generous bill and appealing pastel composition. Spokes, basket weave and feather contours are carefully organized. Both legs connect to the body, though joint definition and pedal contacts are simplified. Consistent wheel placement and a ground shadow establish support; restrained shading gives limited volume.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-sol-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-6-sol","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T22:40:33.293Z","media":{"type":"image","src":"/benchmarks/media/68ded00b3a780e4d.png","alt":"GPT-6 Sol: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:40:33.293Z","quality":{"score":69.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"The raccoon rides the dog's back, and the dog's connected foreleg reaches the steering grip. Expressive faces, a cohesive palette and sweeping wake make an appealing illustration. Vehicle construction and seating remain simplified; broad splash ribbons convey speed more convincingly than physical displacement. The sun has little corresponding reflection or effect on the subjects.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-sol-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-6-sol","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:40:46.874Z","media":{"type":"image","src":"/benchmarks/media/7e13e6811bbfd825.png","alt":"GPT-6 Sol: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:40:46.874Z","quality":{"score":70.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cohesive, appealing beach illustration with framing palms, a warm sunset and a clear reflection path. The turtle, crab, birds and beach ball fulfill the playful brief. Clean silhouettes and restrained colors work well, while simplified anatomy, shallow spatial cues and inconsistent foreground shadows limit depth and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-luna-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-6-luna","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T22:41:02.087Z","media":{"type":"image","src":"/benchmarks/media/2500495802e94c92.png","alt":"GPT-6 Luna: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:02.087Z","quality":{"score":47.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The dark veil, crossed arms, restrained smile and layered landscape clearly evoke Mona Lisa. Clean contours and muted colors create a coherent illustration. Replication is limited by enlarged, outlined eyes, simplified facial volume, broad shoulders and indistinct overlapping hands. Gradients suggest depth, but lack the original's subtle painterly transitions.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.01718219460280624,"y":0.08881576039059919,"scale":1.3297620281214724},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-luna-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-6-luna","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:41:05.959Z","media":{"type":"image","src":"/benchmarks/media/4d21b8a16ac6abd3.png","alt":"GPT-6 Luna: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:05.959Z","quality":{"score":65.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The pelican’s distinctive bill and the coordinated vintage poster palette make an appealing illustration. Circular wheels, orderly spokes and clean lettering strengthen the design. Riding mechanics are less convincing: neither wing reaches the handlebars, pedal contact is ambiguous, and the drivetrain is incomplete. The wheels sit against water without a clear supporting surface. Flat lighting limits volume.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-luna-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-6-luna","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T22:41:11.339Z","media":{"type":"image","src":"/benchmarks/media/a0ac54fb19d85f73.png","alt":"GPT-6 Luna: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:11.339Z","quality":{"score":63.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Expressive faces, warm sunset colors and clean silhouettes create an appealing illustration. The raccoon sits on the dog's back, but the dog's grip and steering linkage remain unclear. The broad hull resembles a speedboat, and flat shading, absent sun reflections and modest water displacement limit depth and motion.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-luna-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-6-luna","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:41:22.333Z","media":{"type":"image","src":"/benchmarks/media/545c592e1a1db127.png","alt":"GPT-6 Luna: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:22.333Z","quality":{"score":59.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"The warm sky, layered sea and framing palms create an appealing beach composition. Animals, birds, toys and reflected sunlight fulfill the brief. Human anatomy remains schematic, and simplified figures offer limited expression. Missing shadows and the dolphin's splashless jump weaken physical grounding, while clean silhouettes and readable small details support the illustration.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"x-ai-grok-4-7-svg-selfie-bc2299c5b9f8e8f6","modelId":"x-ai/grok-4.7","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A dark snail with a geometric spiral shell tests silhouette, curved construction and restrained material shading.","promptHash":"bc2299c5b9f8e8f6","date":"2026-09-29","generatedAt":"2026-09-29T22:41:29.368Z","media":{"type":"image","src":"/benchmarks/media/a16cacb8aac6a275.png","alt":"Grok 4.7: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:29.368Z","quality":{"score":46.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"The bold spiral dominates an appealing, restrained silhouette, with crisp angular turns and subtle charcoal highlights. However, a visible background gap separates the shell from the body, making the shell appear suspended. The connected head and delicate eye stalks read clearly, while shading and small features remain simple.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A dark snail with a geometric spiral shell tests silhouette, curved construction and restrained material shading.","promptSummary":"A dark snail with a geometric spiral shell tests silhouette, curved construction and restrained material shading.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"x-ai-grok-4-7-svg-mona-lisa-58920dfd672bf618","modelId":"x-ai/grok-4.7","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T22:41:36.277Z","media":{"type":"image","src":"/benchmarks/media/e1f5ecc004ed62d7.png","alt":"Grok 4.7: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:36.277Z","quality":{"score":23,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"Dark clothing, overlapping hands and a restrained smile make the reference recognizable. The broad frontal face, elongated neck and oversized head weaken likeness and anatomical proportion; rounded hands lack convincing fingers. Gentle facial shading suggests volume, while clothing and landscape remain flat. Angular veil overlaps and a hard horizontal background boundary weaken the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":2},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":3},{"id":"light","label":"Light & reflections","score":1.5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":2.5},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05085684721636298,"y":0.3321655689361341,"scale":1.9937155332430792},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"x-ai-grok-4-7-svg-pelican-b618cda0f539e818","modelId":"x-ai/grok-4.7","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:41:39.586Z","media":{"type":"image","src":"/benchmarks/media/9d4ab467caf98c96.png","alt":"Grok 4.7: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:39.586Z","quality":{"score":59,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The pelican reads clearly through its long bill, throat pouch and curved neck. Clean pastel shapes and balanced scenery make an appealing illustration. The riding mechanism is less convincing: feet lack clear pedal contact, steering is unresolved, and the frame lacks a coherent drivetrain. Simple shading and sparse feather detail limit depth and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"x-ai-grok-4-7-svg-raccoon-b4cedf42a4a26a4b","modelId":"x-ai/grok-4.7","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T22:41:42.785Z","media":{"type":"image","src":"/benchmarks/media/8b60d8c14b288359.png","alt":"Grok 4.7: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:42.785Z","quality":{"score":56.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon rides clearly on the dog's back, and expressive faces, bright colors and a readable orange hull create a playful composition. Steering is unresolved: the dog's visible paws rest on the deck rather than grasping the controls. Simple body shapes and ribbonlike wake limit construction and movement. The glowing sun produces little corresponding illumination or reflection.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"x-ai-grok-4-7-svg-beach-d030aa6f45fe19e9","modelId":"x-ai/grok-4.7","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:41:48.194Z","media":{"type":"image","src":"/benchmarks/media/7ef283a81e11aac8.png","alt":"Grok 4.7: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:48.194Z","quality":{"score":52.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The warm sunset, blue water and scattered beach activities create a cheerful, readable composition. Animals and equipment use basic silhouettes, with limited anatomical detail and grounding. The sun reflection forms a rigid tapered strip, and foreground objects lack cast shadows. Clean shapes support the illustration, but depth, materials and drawing complexity remain modest.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-svg-selfie-d50ad6124d137ee4","modelId":"minimax/minimax-m3","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small robot with an oversized angular head presents a glowing object. The scene tests character proportions, gesture and contrasting materials.","promptHash":"d50ad6124d137ee4","date":"2026-09-29","generatedAt":"2026-09-29T22:41:49.984Z","media":{"type":"image","src":"/benchmarks/media/8fa8aa3b0656fea1.png","alt":"MiniMax M3: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:49.984Z","quality":{"score":58.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The oversized coral M silhouette dominates clearly, with an expressive charcoal face and appealing miniature body. Limbs connect coherently and the cube reads clearly. The lowered left hand only loosely suggests the requested hip pose. Thin highlights and simple gradients provide limited sculptural volume; the cube's glow and grounding shadow are restrained.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small robot with an oversized angular head presents a glowing object. The scene tests character proportions, gesture and contrasting materials.","promptSummary":"A small robot with an oversized angular head presents a glowing object. The scene tests character proportions, gesture and contrasting materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-svg-mona-lisa-58920dfd672bf618","modelId":"minimax/minimax-m3","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T22:41:55.468Z","media":{"type":"image","src":"/benchmarks/media/ad00cf81f8ae6614.png","alt":"MiniMax M3: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:55.468Z","quality":{"score":19.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered portrait, dark hair and muted palette suggest Mona Lisa, but characteristic likeness and pose are poorly reproduced. The broad oval face lacks her subtle smile; rounded shoulders and a lone mittenlike hand leave the crossed-arm construction unresolved. Warm shading supplies some volume, while blurred landscape bands, generic clothing and indistinct features limit the drawing.","categories":[{"id":"fidelity","label":"Task & likeness","score":2.5},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":2.5},{"id":"light","label":"Light & reflections","score":3},{"id":"detail","label":"Purposeful detail","score":1.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":2}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04312905611593771,"y":0.10911423348597708,"scale":1.4333294145603386},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-svg-pelican-b618cda0f539e818","modelId":"minimax/minimax-m3","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:41:57.323Z","media":{"type":"image","src":"/benchmarks/media/0b931fec3b131a3a.png","alt":"MiniMax M3: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:57.323Z","quality":{"score":66,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The pelican's long bill, pouch and curved neck read clearly, and the warm sunset gives the composition appeal. Bicycle construction weakens the riding action: feet do not convincingly engage the pedals, crank connections are unclear, and body support is unresolved. Clean silhouettes and orderly spokes provide a competent illustrative finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-svg-raccoon-b4cedf42a4a26a4b","modelId":"minimax/minimax-m3","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T22:41:57.660Z","media":{"type":"image","src":"/benchmarks/media/1e7e2c75dd680864.png","alt":"MiniMax M3: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:41:57.660Z","quality":{"score":66.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon sits above the dog's back, and the dog reaches toward the jet ski controls. Expressive faces and a vivid sunset make the scene appealing. Construction is simplified: the dog's paws and steering grip are ambiguous, the hull is shallow, and blurred wake patches give limited water displacement. Sunset glow exceeds the lighting modeled on the riders.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-svg-beach-d030aa6f45fe19e9","modelId":"minimax/minimax-m3","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:42:07.768Z","media":{"type":"image","src":"/benchmarks/media/4f42672ffb0ebc9d.png","alt":"MiniMax M3: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:42:07.768Z","quality":{"score":58,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cheerful, coherent sunset illustration with framing palms, warm gradients and an aligned sun reflection. Animals and beach toys add character, but their construction remains basic. The sailboat appears suspended above the water horizon, while absent cast shadows weaken foreground contact. Clean geometric shapes provide polish without substantial texture or depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-3-8-flash-svg-selfie-174a9113c6779062","modelId":"google/gemini-3.8-flash","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A luminous star-shaped creature tests expressive features, a clear silhouette and smooth colour transitions.","promptHash":"174a9113c6779062","date":"2026-09-29","generatedAt":"2026-09-29T22:42:09.510Z","media":{"type":"image","src":"/benchmarks/media/4846ed211a147ba9.png","alt":"Gemini 3.8 Flash: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:42:09.510Z","quality":{"score":59,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The dominant four-point silhouette, luminous color progression, raised brows and enormous grin closely fulfill the brief. Attached, secondary limbs support the exuberant expression. Clean facial drawing and glossy eyes give strong character, while shallow depth, minimally articulated extremities and visible steps in the outer glow limit refinement.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A luminous star-shaped creature tests expressive features, a clear silhouette and smooth colour transitions.","promptSummary":"A luminous star-shaped creature tests expressive features, a clear silhouette and smooth colour transitions.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-3-8-flash-svg-mona-lisa-58920dfd672bf618","modelId":"google/gemini-3.8-flash","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-29","generatedAt":"2026-09-29T22:42:25.916Z","media":{"type":"image","src":"/benchmarks/media/87c7607365346e35.png","alt":"Gemini 3.8 Flash: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:42:25.916Z","quality":{"score":29.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The dark clothing, overlapping hands and hazy landscape make the Mona Lisa reference recognizable. Replication is limited by the frontal, elongated face, heavy mouth and chin, and poorly resolved hands, wrists and forearm. Soft shading gives some volume, but muddy transitions, blunt landscape shapes and rigid facial modeling weaken the portrait’s expression and painterly character.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":3.5},{"id":"light","label":"Light & reflections","score":3},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":1}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.024677204207361747,"y":0.04654848501348373,"scale":1.17939311871139},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-3-8-flash-svg-pelican-b618cda0f539e818","modelId":"google/gemini-3.8-flash","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:42:30.859Z","media":{"type":"image","src":"/benchmarks/media/20d5148ee65fad01.png","alt":"Gemini 3.8 Flash: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:42:30.859Z","quality":{"score":78.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The pelican’s enormous bill, lively feathers and orange bicycle create a clear, appealing composition. Basket contents, spokes and facial details reward inspection. Both legs connect to the body, but foot-to-pedal placement and wing-to-handlebar contact are ambiguous. Bicycle mechanics and lighting remain schematic, limiting an otherwise polished illustration.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-3-8-flash-svg-raccoon-b4cedf42a4a26a4b","modelId":"google/gemini-3.8-flash","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T22:42:31.136Z","media":{"type":"image","src":"/benchmarks/media/12ca47fc978cf957.png","alt":"Gemini 3.8 Flash: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:42:31.136Z","quality":{"score":68.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Expressive animals, a bold diagonal hull and sweeping spray create an appealing action scene. The raccoon's support on the dog's back is unclear, and the dog lacks a readable paw-to-control grip and connected steering assembly. Clean graphic shapes and lively accessories are stronger than the anatomical construction. Sunset illumination and reflections remain largely decorative.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-3-8-flash-svg-beach-d030aa6f45fe19e9","modelId":"google/gemini-3.8-flash","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:42:38.233Z","media":{"type":"image","src":"/benchmarks/media/d81269c6ae28888d.png","alt":"Gemini 3.8 Flash: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:42:38.233Z","quality":{"score":75.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A vivid purple-and-gold sunset anchors an appealing beach composition with dolphins, birds, a turtle and playful accessories. Palm silhouettes frame the bright reflection effectively. Sparse animal anatomy, wirelike palm foliage, conspicuously regular reflection bands and limited surface texture keep the execution below exceptional.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-27b-svg-selfie-aea705fcb6511459","modelId":"qwen/qwen3.8-27B","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptHash":"aea705fcb6511459","date":"2026-09-29","generatedAt":"2026-09-29T20:24:05.815Z","media":{"type":"image","src":"/benchmarks/media/6be438109080de8e.png","alt":"Qwen3.8-27B: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:24:05.815Z","quality":{"score":77.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cohesive emerald-and-gold illustration with a gentle expression, flowing robes, luminous halo and three clearly ringed orbs. Limbs remain connected, and the asymmetric legs communicate floating. The rounded face reads somewhat bear-like; the raised palm faces outward rather than upward, and chest-hand fingers are crowded. Clean ornamentation exceeds the comparatively simple anatomical modeling and shading.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummary":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-27b-svg-mona-lisa-e7b85d8cd393c26f","modelId":"qwen/qwen3.8-27B","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-09-29","generatedAt":"2026-09-29T20:24:05.816Z","media":{"type":"image","src":"/benchmarks/media/7eb1345ca97bdc24.png","alt":"Qwen3.8-27B: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:24:05.816Z","quality":{"score":47.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered portrait, dark clothing and overlapping hands evoke the Mona Lisa, but replication is weak. The face is excessively elongated, the eyes stare broadly, and the hands reduce to rounded masses with shallow finger marks. Soft shading creates some volume, while the landscape remains indistinct. The precise gold frame is more resolved than the portrait.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05274149094589809,"y":0.10564380033230408,"scale":1.616074402192889},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-27b-svg-pelican-b618cda0f539e818","modelId":"qwen/qwen3.8-27B","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T20:24:05.816Z","media":{"type":"image","src":"/benchmarks/media/8413c5f7e76431a4.png","alt":"Qwen3.8-27B: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:24:05.816Z","quality":{"score":70.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The pelican’s oversized bill, straw hat and flowing scarf create a distinctive focal character against a well-composed coastal sunset. Wheels and frame are readable, but the wings do not reach the steering, overlapping legs obscure separate pedal contacts, and the crank linkage is unclear. Clean vector shapes and restrained shading support an appealing illustration with limited physical depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-27b-svg-raccoon-b4cedf42a4a26a4b","modelId":"qwen/qwen3.8-27B","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T20:24:05.816Z","media":{"type":"image","src":"/benchmarks/media/91229e904e943109.png","alt":"Qwen3.8-27B: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:24:05.816Z","quality":{"score":67.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon clearly rides on the dog's back, and the dog reaches the jet ski's steering grip. Cheerful expressions, a bright palette and sweeping wake make the action readable. Anatomy and machinery remain simplified, with overlapping paws and limited seat definition. Flat lighting, sparse surface detail and broad white splash shapes limit depth and finish. A clipped gray fragment interrupts the upper-left corner.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-27b-svg-beach-d030aa6f45fe19e9","modelId":"qwen/qwen3.8-27B","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T20:24:05.816Z","media":{"type":"image","src":"/benchmarks/media/79fd3f6c2304518b.png","alt":"Qwen3.8-27B: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:24:05.816Z","quality":{"score":68.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-12","rationale":"The sweeping palm, glowing sunset and purple-to-orange palette create an appealing beach composition. Birds, animals and reflections fulfill the brief, with clean silhouettes and readable small accents. Schematic anatomy, repetitive shoreline geometry and limited perspective integration constrain depth. Foreground lighting and uniform reflection bars provide atmosphere but little physical coherence.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-sonnet-5-5-svg-selfie-64223bae34ab729d","modelId":"anthropic/claude-sonnet-5.5","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptHash":"64223bae34ab729d","date":"2026-09-29","generatedAt":"2026-09-29T20:30:05.162Z","media":{"type":"image","src":"/benchmarks/media/8d54c78333e2b22b.png","alt":"Claude Sonnet 5.5: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:30:05.162Z","quality":{"score":79.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The centered coral squid has a friendly, composed expression and a balanced flower-like silhouette. Eight thick arms and two slender tentacles remain visibly connected, with no inverted parts. The mantle is elongated rather than compact, and the arm spacing is uneven. Soft shading gives volume, but dense stippling weakens the requested smooth matte finish; the broad background halo makes depth less precise.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":8},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummary":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-sonnet-5-5-svg-mona-lisa-e7b85d8cd393c26f","modelId":"anthropic/claude-sonnet-5.5","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-09-29","generatedAt":"2026-09-29T20:30:05.162Z","media":{"type":"image","src":"/benchmarks/media/07512d25b662c073.png","alt":"Claude Sonnet 5.5: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:30:05.162Z","quality":{"score":60.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The crossed hands, dark dress, restrained smile and winding landscape clearly recall the Mona Lisa. Replication is limited by the frontal symmetry, wide eyes, elongated neck and flattened fingers. Soft facial shading provides volume, but heavy darkness and pervasive crackle obscure forms. The composition is coherent; the landscape and fabric remain visibly schematic.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.0452734131397336,"y":0.14095434609197272,"scale":1.4993025000567664},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-sonnet-5-5-svg-pelican-b618cda0f539e818","modelId":"anthropic/claude-sonnet-5.5","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T20:30:05.162Z","media":{"type":"image","src":"/benchmarks/media/b87cd2dc13b1cd74.png","alt":"Claude Sonnet 5.5: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:30:05.162Z","quality":{"score":76.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"The pelican is immediately recognizable, with an expressive face, generous bill and appealing cycling pose. Spokes, chain, feathers and scarf provide purposeful detail. The bicycle largely holds together, though overlapping angular legs obscure the far foot’s pedal contact. Clean outlines and balanced composition succeed; flat surfaces and limited directional lighting constrain depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-sonnet-5-5-svg-raccoon-b4cedf42a4a26a4b","modelId":"anthropic/claude-sonnet-5.5","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T20:30:05.162Z","media":{"type":"image","src":"/benchmarks/media/d0cba55b90a4b3d8.png","alt":"Claude Sonnet 5.5: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:30:05.162Z","quality":{"score":76.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon clearly rides on the dog's back while the dog grips the jet ski controls. Expressive faces, vivid colors and a sweeping wake make the scene appealing. Visible limbs connect coherently, but overlapping steering paws and an indistinct control mounting weaken construction. Flat character lighting and simplified water limit depth and material realism.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-sonnet-5-5-svg-beach-d030aa6f45fe19e9","modelId":"anthropic/claude-sonnet-5.5","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T00:21:13.408Z","media":{"type":"image","src":"/benchmarks/media/98f3acd10bb91df6.png","alt":"Claude Sonnet 5.5: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T00:21:13.408Z","quality":{"score":87,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A richly designed beach illustration combines radiant sunset bands, intricate palm fans and playful silhouettes within a cohesive purple, coral and turquoise palette. Carefully varied accessories reward close inspection, while overlapping foreground forms and shoreline layers organize the abundant activity. Dense palm strokes compete with focal subjects; animal anatomy, relative scales and reflected silhouettes remain simplified.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":8.5},{"id":"light","label":"Light & reflections","score":8.5},{"id":"detail","label":"Purposeful detail","score":9.5},{"id":"artistry","label":"Artistry & character","score":9.5},{"id":"effects","label":"Materials & effects","score":8.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-flash-svg-selfie-271fb2af67f0ddcd","modelId":"xiaomi/mimo-v2.6-flash","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small rabbit in an explorer outfit floats with a luminous orb. The scene tests character anatomy, pose and the placement of geometric forms around it.","promptHash":"271fb2af67f0ddcd","date":"2026-09-29","generatedAt":"2026-09-29T22:21:32.369Z","media":{"type":"image","src":"/benchmarks/media/629d36a0889a0936.png","alt":"MiMo-V2.6-Flash: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:21:32.369Z","quality":{"score":67,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"The centered rabbit has an appealing alert expression, a balanced silhouette, and a restrained technological palette. Outfit panels and orbital accents are cleanly resolved. The orb-holding paw has an ambiguous grip, the ears lack distinct angular bends, and the halo's interlocking construction is unclear. Soft shading suggests volume, but materials and spatial depth remain simplified.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small rabbit in an explorer outfit floats with a luminous orb. The scene tests character anatomy, pose and the placement of geometric forms around it.","promptSummary":"A small rabbit in an explorer outfit floats with a luminous orb. The scene tests character anatomy, pose and the placement of geometric forms around it.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-flash-svg-mona-lisa-e7b85d8cd393c26f","modelId":"xiaomi/mimo-v2.6-flash","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-09-29","generatedAt":"2026-09-29T20:44:27.574Z","media":{"type":"image","src":"/benchmarks/media/4064ce9dfb07cf58.png","alt":"MiMo-V2.6-Flash: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:44:27.574Z","quality":{"score":32.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The dark clothing, long hair, folded hands and hazy landscape make the reference recognizable. Replication is limited by the frontal symmetry, elongated face, wide staring eyes and downturned mouth. Both arms connect to the hands, but the overlapping palms and looplike fingers are poorly resolved. Soft facial shading provides some volume; the clothing and landscape remain broad, simplified shapes.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":2.5},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":3},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":2.5},{"id":"effects","label":"Materials & effects","score":3.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04695663175572684,"y":-0.02473107383713599,"scale":1.179393118711389},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-flash-svg-pelican-b618cda0f539e818","modelId":"xiaomi/mimo-v2.6-flash","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-29","generatedAt":"2026-09-29T22:04:15.800Z","media":{"type":"image","src":"/benchmarks/media/fcf26b1b5aba7c3b.png","alt":"MiMo-V2.6-Flash: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:04:15.800Z","quality":{"score":66.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The pelican is immediately recognizable, with an expressive eye and prominent pouch, and the landscape provides clear framing. Clean outlines and spokes support the cheerful illustration. Both feet visibly miss the bicycle's pedal mechanism, weakening the riding action. Simplified machinery, shallow depth and loosely directional shading limit the requested masterpiece quality.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-flash-svg-raccoon-b4cedf42a4a26a4b","modelId":"xiaomi/mimo-v2.6-flash","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-29","generatedAt":"2026-09-29T20:44:30.092Z","media":{"type":"image","src":"/benchmarks/media/a1642a50768bdf22.png","alt":"MiMo-V2.6-Flash: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-29T20:44:30.092Z","quality":{"score":51.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon riding the dog reads clearly, with expressive faces and a cheerful waterscape. The steering action is weak: no paw reaches the handlebar, which meets the dog's muzzle. The vehicle resembles a thin board with a steering post. Overlapping limbs obscure anatomy, while limited shadows, reflections and surface modeling keep the scene flat.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-flash-svg-beach-d030aa6f45fe19e9","modelId":"xiaomi/mimo-v2.6-flash","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-29","generatedAt":"2026-09-29T22:04:14.884Z","media":{"type":"image","src":"/benchmarks/media/cdcaa2b4b5abb435.png","alt":"MiMo-V2.6-Flash: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-29T22:04:14.884Z","quality":{"score":67,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The warm sunset, aligned reflection, framing palms and playful beach animals create an appealing, coherent scene. Small toys and shells remain readable. Drawing is simplified: animals have limited volume, water uses repetitive horizontal strokes, and foreground shadows do not consistently follow the central sun. Clean execution supports a competent illustration, but depth and material rendering remain modest.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-svg-mona-lisa-58920dfd672bf618","modelId":"moonshotai/kimi-k3","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.459Z","media":{"type":"image","src":"/benchmarks/media/c26fe83a8f7333ac.png","alt":"Kimi K3: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.459Z","quality":{"score":42.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered portrait, dark veil, folded hands and misty landscape clearly evoke Mona Lisa. Replication is limited by the elongated neck, simplified facial proportions and oval hands with shallow finger marks. Soft shading gives some volume, but pervasive blur substitutes for controlled painterly transitions and precise detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.0360004951751198,"y":0.07239043315337006,"scale":1.3099644507332462},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-svg-pelican-b618cda0f539e818","modelId":"moonshotai/kimi-k3","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.459Z","media":{"type":"image","src":"/benchmarks/media/5c6b79cbf1e650a8.png","alt":"Kimi K3: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.459Z","quality":{"score":66.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The pelican and bicycle are immediately recognizable, with an expressive bill, flowing scarf and cheerful composition. Both legs remain attached, but pedal connections and wing contact with the handlebar are weakly resolved. Wheel alignment and landscape layers are coherent. Clean outlines support the illustration, while flat surfaces and limited lighting keep it short of masterful execution.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-svg-raccoon-b4cedf42a4a26a4b","modelId":"moonshotai/kimi-k3","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.459Z","media":{"type":"image","src":"/benchmarks/media/e0ccdec24c3e7514.png","alt":"Kimi K3: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.459Z","quality":{"score":71.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"The raccoon rides on the dog's back, and the dog's paw meets a steering assembly connected to the red hull. Expressive faces, a warm sunset and sweeping spray create an appealing composition. Crowded body overlaps obscure some limb connections, and the raccoon's right ear appears slightly separated. Flat shading and stylized reflections limit volume and depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-svg-beach-d030aa6f45fe19e9","modelId":"moonshotai/kimi-k3","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.459Z","media":{"type":"image","src":"/benchmarks/media/58550a0d2679e61c.png","alt":"Kimi K3: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.459Z","quality":{"score":75.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A coordinated sunset palette, expressive silhouettes and varied small details create a lively, appealing beach illustration. The central sun and framing palm and umbrella organize the busy scene, with clean contours preserving readability. Bodies and people remain diagrammatic, depth is shallow, and inverted figure reflections cross matte-looking sand without convincing surface lighting. Materials and movement are mostly conveyed through flat shapes.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"z-ai-glm-5-3-svg-mona-lisa-58920dfd672bf618","modelId":"z-ai/glm-5.3","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.552Z","media":{"type":"image","src":"/benchmarks/media/3282985e6d682b68.png","alt":"GLM-5.3: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.552Z","quality":{"score":30,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"Recognizable Mona Lisa motifs include the dark veil, restrained smile, folded hands and muted landscape. Clean silhouettes and balanced framing support the portrait, but the frontal symmetry, elongated face, minimally articulated hands and broad background bands make the replication basic. Facial shading offers some volume; clothing and landscape remain mostly flat.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":3},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":2.5},{"id":"effects","label":"Materials & effects","score":3.5},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05526821096442614,"y":0.26487173945070097,"scale":1.741940984774073},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"z-ai-glm-5-3-svg-pelican-b618cda0f539e818","modelId":"z-ai/glm-5.3","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.552Z","media":{"type":"image","src":"/benchmarks/media/b99e8995a16fde8d.png","alt":"GLM-5.3: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.552Z","quality":{"score":56.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The expressive pelican, broad bill and clean coastal composition make the subject immediately readable. Both legs connect to the body and reach pedals, but the front wheel lacks a connected fork and the wings do not engage visible handlebars. Crisp spokes and feather lines add interest; flat lighting, absent ground shadows and a fish floating above the horizon weaken physical coherence.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"z-ai-glm-5-3-svg-raccoon-b4cedf42a4a26a4b","modelId":"z-ai/glm-5.3","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.552Z","media":{"type":"image","src":"/benchmarks/media/1743b5220fff46ea.png","alt":"GLM-5.3: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.552Z","quality":{"score":54.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon rides on the dog's back, and the dog reaches toward visible handlebars. Expressive faces and clean colors make the action readable. Overlapping paws and hindquarters weaken anatomical clarity, while the vehicle resembles a flat board with a thin steering post. The wake suggests motion, but water contact, shadows and reflections receive little development.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-astra-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-6-astra","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T01:39:52.826Z","media":{"type":"image","src":"/benchmarks/media/45e7730bc430db1d.png","alt":"GPT-6 Astra: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:52.826Z","quality":{"score":66.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"The crossed hands, restrained smile, dark clothing and hazy landscape clearly recall the Mona Lisa. Replication is limited by the frontal, elongated face, schematic eyes and stiff fingers. Soft shading gives basic volume, but broad garment shapes and angular scenery lack the original’s nuanced depth and painterly transitions. The frame and composition are controlled; surface grain adds texture without resolving the simplified drawing.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.042334746078057595,"y":0.06525087361717094,"scale":1.3498588075760019},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-astra-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-6-astra","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T01:40:04.067Z","media":{"type":"image","src":"/benchmarks/media/c5417b121961a258.png","alt":"GPT-6 Astra: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:40:04.067Z","quality":{"score":76,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"The pelican's long bill, flowing scarf and warm coastal palette create a charming, balanced illustration. Wheels, spokes and frame are largely coherent, but neither wing reaches the handlebars, and the rear leg's pedal contact is unclear. Flat lighting and simplified surfaces limit depth despite clean outlines and attractive feather detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-astra-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-6-astra","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T01:42:11.403Z","media":{"type":"image","src":"/benchmarks/media/dd9d134299d604c5.png","alt":"GPT-6 Astra: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:42:11.403Z","quality":{"score":78,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Expressive animals, a balanced sunset composition and sweeping wake make an appealing illustration. The raccoon rides the dog's back, and the dog's connected foreleg reaches the controls. The compact steering assembly and layered paws are less clearly constructed. Lighting remains mostly flat, with little shadow or sun reflection. Clean contours and purposeful details support strong illustrative execution.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":8},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-astra-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-6-astra","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T01:42:16.633Z","media":{"type":"image","src":"/benchmarks/media/2c9850c70ab79b4c.png","alt":"GPT-6 Astra: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:42:16.633Z","quality":{"score":80.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"A richly composed sunset beach illustration with cohesive peach, mauve and teal colors. Sweeping palm fronds, a scalloped umbrella and curved shorelines frame the sun and its aligned reflection. Distinct animal silhouettes and carefully distributed beach details reward inspection. Simplified human and dog anatomy, slight dolphin splash cues and inconsistent equipment shadows limit physical conviction.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-opus-5-5-svg-selfie-64223bae34ab729d","modelId":"anthropic/claude-opus-5.5","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptHash":"64223bae34ab729d","date":"2026-09-30","generatedAt":"2026-09-30T01:46:12.074Z","media":{"type":"image","src":"/benchmarks/media/27adf2c7f970ba94.png","alt":"Claude Opus 5.5: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:46:12.074Z","quality":{"score":75.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The centered coral squid has a friendly, restrained expression and a balanced radial silhouette within a cream arch. Smooth curves and soft shading give it an appealing matte finish. The mantle is more elongated than compact, the upper arms are visually ambiguous, and the two hanging tentacles read somewhat like legs. Clean execution supports a strong stylized illustration.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummary":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-opus-5-5-svg-mona-lisa-58920dfd672bf618","modelId":"anthropic/claude-opus-5.5","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T01:42:58.567Z","media":{"type":"image","src":"/benchmarks/media/7fd07292388572f2.png","alt":"Claude Opus 5.5: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:42:58.567Z","quality":{"score":62.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The portrait preserves the recognizable composition, dark veil, folded hands and distant landscape. Replication is limited by the elongated face and neck, conspicuous eyes and simplified smile. Hands remain connected but their rounded fingers and wrists lack convincing anatomy. Restrained colors create cohesion; shallow modeling and pervasive mottling weaken volume and painterly subtlety.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.03691937129169585,"y":0.08755011578607834,"scale":1.5449630589513361},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-opus-5-5-svg-pelican-b618cda0f539e818","modelId":"anthropic/claude-opus-5.5","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T01:44:09.522Z","media":{"type":"image","src":"/benchmarks/media/c4f0783a8524a30d.png","alt":"Claude Opus 5.5: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:44:09.522Z","quality":{"score":76,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The recognizable pelican, orderly bicycle wheels and harmonious sunset create an appealing illustration. Both legs connect to the body and reach pedals; no clear detached or inverted parts appear. Steering contact and seating support are less convincing, while the drivetrain remains schematic. Crisp outlines give a clean finish, but limited shading and flat surfaces leave the scene with modest depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-opus-5-5-svg-raccoon-b4cedf42a4a26a4b","modelId":"anthropic/claude-opus-5.5","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T01:45:08.147Z","media":{"type":"image","src":"/benchmarks/media/7e0a20b80d132e35.png","alt":"Claude Opus 5.5: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:45:08.147Z","quality":{"score":76,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The masked raccoon clearly rides the dog's back, and the dog's forward paw reaches the steering grip. Expressive faces, vivid colors and sweeping wakes create an appealing illustration. Connected limbs remain schematic, as does the small steering assembly. The broad hull and flat foam weaken water contact; the glowing sun lacks a corresponding reflection or convincing cast shadows.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-opus-5-5-svg-beach-d030aa6f45fe19e9","modelId":"anthropic/claude-opus-5.5","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T01:45:35.304Z","media":{"type":"image","src":"/benchmarks/media/7efbaea68b520243.png","alt":"Claude Opus 5.5: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:45:35.304Z","quality":{"score":77,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A lively, well-composed beach illustration with rich sunset colors, recognizable animals and appealing accessories. The sun reflection establishes a strong central focus. Clean silhouettes and small details reward inspection, though simplified animal construction, uneven contact shadows and mostly flat materials limit depth and realism.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-fable-5-1-svg-selfie-64223bae34ab729d","modelId":"anthropic/claude-fable-5.1","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptHash":"64223bae34ab729d","date":"2026-09-30","generatedAt":"2026-09-30T01:47:58.884Z","media":{"type":"image","src":"/benchmarks/media/ea37603e33e0b232.png","alt":"Claude Fable 5.1: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:47:58.884Z","quality":{"score":68.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The coral palette, friendly face, matte shading and cream arch create a coherent emblem. However, the mantle is elongated rather than compact, and the appendages spread mostly sideways and downward instead of forming an evenly spaced, flower-like starburst. Two thin hanging appendages weaken the thick-arm motif. Clean contours and balanced framing support the simple design.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummary":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-fable-5-1-svg-mona-lisa-58920dfd672bf618","modelId":"anthropic/claude-fable-5.1","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T01:47:58.884Z","media":{"type":"image","src":"/benchmarks/media/d62e37155545dfbc.png","alt":"Claude Fable 5.1: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:47:58.884Z","quality":{"score":57.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The dark clothing, crossed hands and hazy landscape establish the Mona Lisa composition. Replication is limited by the frontal, elongated face, wide eyes and simplified smile. Rounded, parallel fingers and unclear wrist transitions weaken the hands. Warm facial shading gives some volume, but broad garment shapes and scattered crack marks substitute for convincing folds and painterly detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.03154097782210446,"y":0.057703405914141,"scale":1.1618342427282822},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-fable-5-1-svg-pelican-b618cda0f539e818","modelId":"anthropic/claude-fable-5.1","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T01:47:58.884Z","media":{"type":"image","src":"/benchmarks/media/72ac6a49fe15b1ad.png","alt":"Claude Fable 5.1: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:47:58.884Z","quality":{"score":72.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The pelican reads clearly, with an expressive face, prominent bill and appealing composition. Both legs connect to the body, but the far foot's pedal contact and the wing's steering grip remain unclear. Wheel alignment and background layers are coherent. Clean contours and feather lines give polish, while schematic machinery, limited surface texture and weak directional lighting constrain the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-fable-5-1-svg-raccoon-b4cedf42a4a26a4b","modelId":"anthropic/claude-fable-5.1","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T01:47:58.884Z","media":{"type":"image","src":"/benchmarks/media/37e9e4be6ff352fe.png","alt":"Claude Fable 5.1: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:47:58.884Z","quality":{"score":70,"rubricId":"svg-visual-v2","reviewedAt":"2026-09-30","rationale":"The raccoon rides on the dog's back, and the dog's extended foreleg reaches the jet ski controls. Expressive faces, a clear silhouette and sweeping spray make the action readable. Limb and handlebar overlaps weaken construction; broad shading provides limited volume, and the glowing sun has little corresponding reflection on the water. Clean vector shapes support an appealing but simplified illustration.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-fable-5-1-svg-beach-d030aa6f45fe19e9","modelId":"anthropic/claude-fable-5.1","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T01:50:08.762Z","media":{"type":"image","src":"/benchmarks/media/546bbd9bc8d79386.png","alt":"Claude Fable 5.1: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:50:08.762Z","quality":{"score":73.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The colorful sunset, layered shoreline, animals, kite and beach toys fulfill the lively beach brief. Clean silhouettes and a balanced composition give the illustration character. Construction is simplified: the umbrella resembles a segmented fan, dolphins have weak water interaction, and the sun’s reflection forms a rigid widening wedge. Flat surfaces and sparse contact shadows limit depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-svg-selfie-51df87a90bcca7ff","modelId":"moonshotai/kimi-k3","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptHash":"51df87a90bcca7ff","date":"2026-09-30","generatedAt":"2026-09-30T01:39:36.459Z","media":{"type":"image","src":"/benchmarks/media/0329da4141ac86e1.png","alt":"Kimi K3: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:39:36.459Z","quality":{"score":78.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The indigo creature and warm reflected counterpart create a balanced, contemplative composition. Equations, constellations and book emblems give the transformation meaning. Clean silhouettes and controlled glow support the scene, but minimally articulated legs, shallow spatial cues and smooth, generalized surfaces limit its anatomical and material richness.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummary":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"z-ai-glm-5-3-svg-selfie-51df87a90bcca7ff","modelId":"z-ai/glm-5.3","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptHash":"51df87a90bcca7ff","date":"2026-09-30","generatedAt":"2026-09-30T01:57:02.634Z","media":{"type":"image","src":"/benchmarks/media/47cd296284db3645.png","alt":"GLM-5.3: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:57:02.634Z","quality":{"score":76.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A luminous organic creature meets a geometric reflection in a carefully framed oval mirror. Shared glowing cores and fingertip contact make the introspective symbolism clear. Gold, cream and blue establish a controlled hierarchy. The creature’s lower anatomy and shoulder transition remain loosely defined, while ground contact and reflected depth are mostly symbolic. Elegant vector illustration rather than photographic rendering.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummary":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"z-ai-glm-5-3-svg-beach-d030aa6f45fe19e9","modelId":"z-ai/glm-5.3","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T01:57:02.634Z","media":{"type":"image","src":"/benchmarks/media/48fb10f86f22073f.png","alt":"GLM-5.3: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T01:57:02.634Z","quality":{"score":75.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cohesive, colorful beach illustration with an inviting sunset, readable animals and playful foreground objects. Crisp silhouettes and small accessories reward inspection. Depth remains shallow, airborne dolphins lack convincing launch splashes, and several objects lack grounding shadows. The broad dark shape across the sun weakens the focal hierarchy.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-1-3-svg-selfie-51df87a90bcca7ff","modelId":"meta/muse-spark-1.3","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptHash":"51df87a90bcca7ff","date":"2026-09-30","generatedAt":"2026-09-30T02:19:58.155Z","media":{"type":"image","src":"/benchmarks/media/8eb1d67a0662309d.png","alt":"Muse Spark 1.3: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T02:19:58.155Z","quality":{"score":63.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The luminous creature and geometric mirror counterpart communicate self-recognition through a coherent palette and expressive eyes. The transformed reflection is purposeful. Simple oval anatomy, ambiguous reaching-arm construction, shallow spatial staging and limited illumination of nearby surfaces keep the illustration below exceptional execution.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummary":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-1-3-svg-mona-lisa-58920dfd672bf618","modelId":"meta/muse-spark-1.3","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T02:19:58.155Z","media":{"type":"image","src":"/benchmarks/media/05681dec424f2dcc.png","alt":"Muse Spark 1.3: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T02:19:58.155Z","quality":{"score":15,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"Dark hair, a restrained smile and overlapping hands suggest Mona Lisa, but replication is weak. The face is markedly elongated, the hands are oval pads with faint finger strokes, and the atmospheric landscape is absent. Soft gradients provide basic volume, while the frontal symmetry, shallow space and sparse clothing detail limit the portrait’s drawing quality.","categories":[{"id":"fidelity","label":"Task & likeness","score":1},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":1.5},{"id":"light","label":"Light & reflections","score":1.5},{"id":"detail","label":"Purposeful detail","score":1},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":1.5},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.09314298771447593,"y":0.6212440990928104,"scale":2.6912344723492607},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-1-3-svg-pelican-b618cda0f539e818","modelId":"meta/muse-spark-1.3","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T02:19:58.155Z","media":{"type":"image","src":"/benchmarks/media/f851c1f64ab4d2ed.png","alt":"Muse Spark 1.3: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T02:19:58.155Z","quality":{"score":66.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The pelican and bicycle are immediately recognizable, with an expressive face, bold bill and cheerful composition. Both legs connect to the body and reach pedal-like supports, but articulation, saddle support and wing contact with the handlebar remain unclear. Wheel alignment and layering are coherent. Simple landscape shapes, generic shadows and limited material variation constrain the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-1-3-svg-raccoon-b4cedf42a4a26a4b","modelId":"meta/muse-spark-1.3","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T02:19:58.155Z","media":{"type":"image","src":"/benchmarks/media/a1bf783a711f5123.png","alt":"Muse Spark 1.3: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T02:19:58.155Z","quality":{"score":66,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The raccoon clearly rides on the dog's back, and the dog's paws reach the jet ski controls. Expressive faces and bright colors make the scene appealing. Simplified joints, crowded steering geometry and shallow hull perspective limit construction. The sun has no corresponding water reflection, and the splash shapes convey speed more clearly than convincing displacement.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-1-3-svg-beach-d030aa6f45fe19e9","modelId":"meta/muse-spark-1.3","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T02:19:58.155Z","media":{"type":"image","src":"/benchmarks/media/d28bdeaa854331fc.png","alt":"Muse Spark 1.3: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T02:19:58.155Z","quality":{"score":44.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The warm sunset, framing palms and colorful beach accessories create an appealing scene with the requested animals, birds and reflections. Construction remains schematic: the raised hand misses the kite string, and a doglike silhouette hangs above the water without clear support. The sun reflection establishes a focal path, but shadows and materials lack depth. An unexplained horizontal bar crosses the sun.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":3},{"id":"integrity","label":"Visual integrity","score":3.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-1-sol-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-6.1-sol","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.855361+00:00","media":{"type":"image","src":"/benchmarks/media/e7a4a2f9207d603e.png","alt":"GPT-6.1 Sol: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.855361+00:00","quality":{"score":76.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Six outward-facing heads and evenly arranged loops make the pale interlocking knot immediately readable. Uniform widths, angular inner bends and fine scales preserve its ornamental geometry. The bodies feel more like textured tubes than living muscle: shading is subdued, crossing contact shadows are weak, and visible curved seams interrupt surface continuity. Small heads have readable eyes and jaws but limited anatomical depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-sol-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-5.6-sol","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.856879+00:00","media":{"type":"image","src":"/benchmarks/media/6cc8f64a16fdc84f.png","alt":"GPT-5.6 Sol: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.856879+00:00","quality":{"score":61.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Six evenly spaced lobes, outward-facing heads and a clear central hexagon establish the requested knot. The ivory silhouette is controlled, but broad crossing bands resemble rigid strips, shading provides limited muscular volume, and repetitive scale lines extend across black loop openings. Small oval heads have readable eyes and mouths but limited anatomical character or moist surface detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-terra-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-5.6-terra","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.865450+00:00","media":{"type":"image","src":"/benchmarks/media/fe4d0248eda84806.png","alt":"GPT-5.6 Terra: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.865450+00:00","quality":{"score":62.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Six ivory heads and rounded lobes create a striking radial emblem around a clear hexagonal opening. Soft shading gives the bodies volume, but abrupt rectangular and diagonal band endings make the continuous alternating weave ambiguous. Repetitive scale lines and simplified faces convey an ornamental serpent more convincingly than living anatomy.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-sol-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-6-sol","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.856879+00:00","media":{"type":"image","src":"/benchmarks/media/951381ba36e71d07.png","alt":"GPT-6 Sol: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.856879+00:00","quality":{"score":65.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Six evenly spaced lobes, small outward-facing heads and a regular hexagonal opening make the motif immediately readable. The central angular band merges several intersections, weakening the alternating weave. Pale shading gives modest volume, but scales are sparse and repetitive, crossing shadows are weak, and the heads have limited anatomical definition.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-luna-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-6-luna","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.856879+00:00","media":{"type":"image","src":"/benchmarks/media/9757c140cad8bfb0.png","alt":"GPT-6 Luna: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.856879+00:00","quality":{"score":45.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Sixfold symmetry, pale coloring and six outward-facing heads establish the motif. However, every neck floats clear of the knot. Squared outer sections replace the required semicircular lobes, and the large stepped opening lacks a regular hexagonal contour. Repeated scales and soft shading suggest skin, but translucent crossings and doubled edges weaken the living, interwoven construction.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-astra-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-6-astra","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.855361+00:00","media":{"type":"image","src":"/benchmarks/media/5727401e8bbcdd5e.png","alt":"GPT-6 Astra: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.855361+00:00","quality":{"score":80.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"The sixfold silhouette, pale coloring, outward heads and central hexagon closely match the brief. Fine scales and soft shading support rounded volume. However, several crossings have conspicuous rectangular seams and weak contact shadows, making the weave feel assembled. Small heads have limited anatomical definition, and the repetitive texture conveys less organic life than requested.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":8},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-luna-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-5.6-luna","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.866449+00:00","media":{"type":"image","src":"/benchmarks/media/6b244240394a8559.png","alt":"GPT-5.6 Luna: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.866449+00:00","quality":{"score":59.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Six pale, outward-facing heads and six rounded lobes make the knot recognizable. The central hexagonal opening remains clear. However, broad gray crossing panels flatten the weave, several inner joins resemble capped strips, and sparse oversized scale marks leave the bodies looking more like decorated ribbons than living snakes.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-luna-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-5.6-luna","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.572024+00:00","media":{"type":"image","src":"/benchmarks/media/e4a4be4ee2f6c5ad.png","alt":"GPT-5.6 Luna: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.572024+00:00","quality":{"score":39.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The central portrait, dark clothing, crossed hands and muted landscape clearly evoke Mona Lisa. Replication is limited by the elongated face, conspicuous eyes and simplified smile. Overlapping fingers have uncertain anatomy, while broad gradients and heavy outlines flatten the portrait. The composition and restrained palette remain coherent.","categories":[{"id":"fidelity","label":"Task & likeness","score":3.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":3.5},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05512596360757882,"y":0.14728128451752615,"scale":1.690458848379089},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-luna-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-5.6-luna","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.572024+00:00","media":{"type":"image","src":"/benchmarks/media/6b7d52e1f8558a0e.png","alt":"GPT-5.6 Luna: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.572024+00:00","quality":{"score":66.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"A recognizable pelican cyclist with an expressive profile, balanced composition and controlled outlines. Head, neck and torso connect coherently. The bicycle wheels and frame are convincing, but overlapping feet obscure pedal contact, the drivetrain lacks a clear chain, and the wing's handlebar grip is schematic. Atmospheric color succeeds more strongly than directional lighting or material depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-luna-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-5.6-luna","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.572024+00:00","media":{"type":"image","src":"/benchmarks/media/acaa1a4e38877c8d.png","alt":"GPT-5.6 Luna: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.572024+00:00","quality":{"score":58.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Expressive animals, a striped tail and clean outlines make the scene appealing and readable. The raccoon sits convincingly on the dog's back, but the overlapping steering paws are ambiguous and the control stem stops above the hull. Broad wake shapes suggest motion; the glowing sun lacks a corresponding water reflection.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-luna-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-5.6-luna","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T18:05:11.110416+00:00","media":{"type":"image","src":"/benchmarks/media/99993f30db761ea8.png","alt":"GPT-5.6 Luna: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:05:11.110416+00:00","quality":{"score":47.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"The warm sunset, palm silhouette and varied beach activities create an appealing, readable composition. Animals, birds and a broad sun reflection satisfy the brief. Drawing remains elementary: angular airborne sea animals lack splash or convincing motion, the crab's claws are awkwardly constructed, and repeated reflection bands suggest a stepped platform. Clean outlines and pleasant colors exceed the anatomical and spatial sophistication.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":3.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-5-svg-selfie-efa9c627eda55ad5","modelId":"openai/gpt-5.5","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptHash":"efa9c627eda55ad5","date":"2026-09-30","generatedAt":"2026-09-30T23:46:58.866449+00:00","media":{"type":"image","src":"/benchmarks/media/ff5cba19e8a83b62.png","alt":"GPT-5.5: Selfie SVG benchmark"},"firstGeneratedAt":"2026-09-30T23:46:58.866449+00:00","quality":{"score":41.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"Six pale, scaled lobes form a balanced radial arrangement, but they remain separate capsules rather than an interlocking serpent knot. There are no crossings, alternating overlaps, or crisp inner bends; the central hexagon is an isolated outline. Small outward-facing heads sit within the lobes without integrated necks. Soft shading and tidy contours provide some finish, while repetitive oval scales and cartoon faces limit organic credibility.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummary":"A pale serpent woven into an interlocking knot tests connected anatomy, repeated structure and readable overlaps. Organic surface detail must work with the overall geometry.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-5-svg-mona-lisa-58920dfd672bf618","modelId":"openai/gpt-5.5","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-09-30","generatedAt":"2026-09-30T18:06:59.004500+00:00","media":{"type":"image","src":"/benchmarks/media/c55c6431ace949f7.png","alt":"GPT-5.5: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:06:59.004500+00:00","quality":{"score":17.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The veil, restrained smile, crossed hands and muted landscape establish Mona Lisa references. The oversized frontal face and schematic features weaken likeness and expression, while fingerless tapered hand shapes leave anatomy unresolved. Clean contours and balanced framing provide coherence, but sparse garment detail, broad landscape bands and shallow shading limit the portrait's sophistication.","categories":[{"id":"fidelity","label":"Task & likeness","score":1.5},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":1.5},{"id":"light","label":"Light & reflections","score":2},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":1.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.06176508688609695,"y":0.4152764830946679,"scale":2.2818807653293027},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-5-svg-pelican-b618cda0f539e818","modelId":"openai/gpt-5.5","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-09-30","generatedAt":"2026-09-30T18:07:52.383050+00:00","media":{"type":"image","src":"/benchmarks/media/399ca795c8854776.png","alt":"GPT-5.5: Pelican SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:07:52.383050+00:00","quality":{"score":60,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The oversized bill, bright bicycle and sunset create an appealing, recognizable cartoon. Riding mechanics are less convincing: the wing has no clear steering contact, leg attachments are uneven, and feet do not visibly engage pedals. A dark notch complicates the rear neck junction. Clean contours and orderly spokes provide polish, while simplified anatomy and lighting limit depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-5-svg-raccoon-b4cedf42a4a26a4b","modelId":"openai/gpt-5.5","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-09-30","generatedAt":"2026-09-30T18:08:09.157449+00:00","media":{"type":"image","src":"/benchmarks/media/93b9a2b50335368b.png","alt":"GPT-5.5: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:08:09.157449+00:00","quality":{"score":37.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-24","rationale":"The raccoon's masked face, bright hull and sweeping wake make the scene readable and lively. The dog is much less resolved: its head is ambiguous, overlapping brown forms obscure its anatomy, and its paws do not visibly grip the controls. Smooth shapes and cheerful colors provide appeal, but weak steering construction and limited depth keep the drawing uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":3.5},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-5-svg-beach-d030aa6f45fe19e9","modelId":"openai/gpt-5.5","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-09-30","generatedAt":"2026-09-30T18:08:51.261347+00:00","media":{"type":"image","src":"/benchmarks/media/bff497b7a18d62e9.png","alt":"GPT-5.5: Beach SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:08:51.261347+00:00","quality":{"score":34.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The sunset, palms, birds, playful animals and ball clearly establish the requested beach scene. Warm sky gradients and clean silhouettes make it appealing. Animal construction is underdeveloped: torso and limb connections are unclear, especially in the round-eared foreground figure. The dolphin lacks a convincing splash, and broad reflection bands suggest sunset light without resolving its behavior on water.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":2.5},{"id":"space","label":"Space & placement","score":3},{"id":"light","label":"Light & reflections","score":3},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":3},{"id":"integrity","label":"Visual integrity","score":3.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"fledge-alpha-svg-raccoon-b4cedf42a4a26a4b","modelId":"fledge-alpha","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-02","generatedAt":"2026-10-02T01:17:31.399Z","media":{"type":"image","src":"/benchmarks/media/4e337a96164be39f.png","alt":"Fledge Alpha: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-02T01:17:31.399Z","quality":{"score":65.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"An appealing sunset illustration with recognizable characters and a clear riding action. Color, expression and silhouette are stronger than anatomical construction: overlapping paws obscure support, the dog's steering limb is simplified, and the vehicle controls lack definition. The wake conveys movement, while lighting and water reflections remain decorative.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"fledge-alpha-svg-selfie-1d4e1ecb8e36c2ae","modelId":"fledge-alpha","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An eager young seabird running toward the viewer tests expressive anatomy, balance and a convincing forward-moving pose.","promptHash":"1d4e1ecb8e36c2ae","date":"2026-10-02","generatedAt":"2026-10-02T01:18:37.649Z","media":{"type":"image","src":"/benchmarks/media/911fa9162b7b7bd5.png","alt":"Fledge Alpha: Selfie SVG benchmark","thumbnail":"/assets/homepage/thumbnails/17e4aa7e43dd5d4ffd227647.png"},"firstGeneratedAt":"2026-10-02T01:18:37.649Z","quality":{"score":61.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A coherent, appealing mascot with expressive eyes, a cheerful beak and a readable swept-back silhouette. Its larger forward foot and trailing feathers suggest a running stride, but the upright body, stiff legs and minimally articulated feet limit convincing motion. Clean flat shapes suit the brief, while uneven feather junctions and incomplete cream accents reduce polish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An eager young seabird running toward the viewer tests expressive anatomy, balance and a convincing forward-moving pose.","promptSummary":"An eager young seabird running toward the viewer tests expressive anatomy, balance and a convincing forward-moving pose.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"fledge-alpha-svg-mona-lisa-634b0d4018106ba9","modelId":"fledge-alpha","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"634b0d4018106ba9","date":"2026-10-02","generatedAt":"2026-10-02T01:39:34.649Z","media":{"type":"image","src":"/benchmarks/media/29e924c96cef17ba.png","alt":"Fledge Alpha: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-02T01:31:57.986Z","quality":{"score":28.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A recognizable but basic Mona Lisa reinterpretation with a coherent silhouette and subdued palette. Curtain-like hair obscures much of the face, while elongated facial construction and oversized, minimally articulated hands substantially weaken replication. Soft shading supplies modest volume, but anatomy, landscape depth and purposeful detail remain underdeveloped.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":2.5},{"id":"light","label":"Light & reflections","score":2.5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":2.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.055040566786310724,"y":0.035125859019776356,"scale":1.3909681284637794},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"fledge-alpha-svg-beach-b42f1ce85bb99124","modelId":"fledge-alpha","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"b42f1ce85bb99124","date":"2026-10-02","generatedAt":"2026-10-02T01:41:37.364Z","media":{"type":"image","src":"/benchmarks/media/1989853c546f79dd.png","alt":"Fledge Alpha: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-02T01:39:46.982Z","quality":{"score":43.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cheerful, recognizable sunset beach with a cohesive warm palette and clear accessories. Simple construction, uncertain palm support, an ambiguous upper boundary and schematic reflections limit depth and finish. The scene communicates its subject successfully but lacks the developed forms, lighting and detail of exceptional illustration.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"fledge-alpha-svg-pelican-23730eba3763743d","modelId":"fledge-alpha","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"23730eba3763743d","date":"2026-10-02","generatedAt":"2026-10-02T01:32:23.145Z","media":{"type":"image","src":"/benchmarks/media/4d456586536a0b4a.png","alt":"Fledge Alpha: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-02T01:32:23.145Z","quality":{"score":60.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A clear, cheerful illustration with a recognizable pelican, coherent bicycle silhouette and tidy composition. Simplified leg articulation, steering contact, mechanical detail and lighting keep it at a competent illustrative level.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-pro-0813-svg-selfie-c14ada77a4644cf8","modelId":"deepseek/deepseek-v4-pro-0813","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptHash":"c14ada77a4644cf8","date":"2026-10-04","generatedAt":"2026-10-04T15:06:56.232Z","media":{"type":"image","src":"/benchmarks/media/1afc3fd43cd5b7c4.png","alt":"DeepSeek-V4-Pro-0813: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:06:56.232Z","quality":{"score":42.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The whale’s warm eye, lavender assistant and tilted shell convey appealing character. However, the shrimp and reciprocal inspection are unresolved, the large bubble and upper tail are cropped, and the flipper’s grip is unclear. Broad competing color fields and conspicuous stepped edges weaken the flowing composition. Gradients suggest volume, but lighting and translucent materials remain uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":2.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptSummary":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-svg-selfie-c14ada77a4644cf8","modelId":"deepseek/deepseek-v4.1-flash","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptHash":"c14ada77a4644cf8","date":"2026-10-04","generatedAt":"2026-10-04T15:06:56.343Z","media":{"type":"image","src":"/benchmarks/media/9acf86120b34b77d.png","alt":"DeepSeek-V4.1-Flash: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:06:56.343Z","quality":{"score":67.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"Expressive faces, cobalt coloring and a flowing diagonal composition make the mutual inspection charming. Execution is uneven: the long pointed flipper lacks a convincing grip, the enlarged shrimp face appears detached from its body, and the large lens is oval. Flat anatomical shading and an unexplained purple ringed patch beneath the whale weaken the requested polished depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptSummary":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"union-alpha-svg-selfie-afe28cfb6bda67b6","modelId":"union-alpha","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An interwoven metallic emblem tests ribbon geometry, overlapping forms and the appearance of reflective metal against a dark background.","promptHash":"afe28cfb6bda67b6","date":"2026-10-01","generatedAt":"2026-10-01T16:55:26.171Z","media":{"type":"image","src":"/benchmarks/media/6a03c1fdcff24900.png","alt":"Union Alpha: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-01T16:55:26.171Z","quality":{"score":76.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A strong futuristic emblem with clear hierarchy, controlled negative space and distinctive typography. Silver and gunmetal layers give the sweeping ribbons convincing graphic depth, while red accents remain restrained. Ambiguous concealed connections and occasionally abrupt highlight bands limit the construction and material realism.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An interwoven metallic emblem tests ribbon geometry, overlapping forms and the appearance of reflective metal against a dark background.","promptSummary":"An interwoven metallic emblem tests ribbon geometry, overlapping forms and the appearance of reflective metal against a dark background.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-flash-svg-selfie-aea705fcb6511459","modelId":"qwen/qwen3.8-flash","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptHash":"aea705fcb6511459","date":"2026-10-02","generatedAt":"2026-10-02T05:05:28.568Z","media":{"type":"image","src":"/benchmarks/media/ab435aaafdac4902.png","alt":"Qwen3.8-Flash: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-02T05:05:28.568Z","quality":{"score":76.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A coherent, appealing fantasy illustration with disciplined framing, luminous teal accents and consistent ornament. Most requested elements are clearly present. The bear-like facial proportions, lowered resting hand, obscured bent-leg construction and limited fabric modeling constrain likeness and anatomical clarity; glow effects are attractive but weakly integrated with surrounding surfaces.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummary":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"tencent-hy4-preview-svg-selfie-5d476adbc3300ca6","modelId":"tencent/hy4-preview","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A young blue dragon floating among clouds tests curved anatomy, expressive features and the contrast between soft clouds and glossy surfaces.","promptHash":"5d476adbc3300ca6","date":"2026-10-04","generatedAt":"2026-10-04T15:51:38.600Z","media":{"type":"image","src":"/benchmarks/media/0946f7a4f5465358.png","alt":"Hy4 preview: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:51:38.600Z","quality":{"score":60.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A cohesive, appealing illustration with a serene face, luminous palette and complete framing. The separated head and torso weaken anatomical coherence, while crowded wing attachments and angular sash folds limit the requested suppleness. Clean execution and decorative detail exceed the clarity of the underlying construction.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A young blue dragon floating among clouds tests curved anatomy, expressive features and the contrast between soft clouds and glossy surfaces.","promptSummary":"A young blue dragon floating among clouds tests curved anatomy, expressive features and the contrast between soft clouds and glossy surfaces.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"tencent-hy4-preview-svg-mona-lisa-e7b85d8cd393c26f","modelId":"tencent/hy4-preview","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-04","generatedAt":"2026-10-04T15:51:38.600Z","media":{"type":"image","src":"/benchmarks/media/799b3e4b1a080888.png","alt":"Hy4 preview: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:51:38.600Z","quality":{"score":23.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"A recognizable but basic Mona Lisa interpretation with a cohesive earthy palette and atmospheric hills. The rigid neck, generic face and unresolved folded-hand pose limit replication and anatomy. Large silhouettes and layer boundaries are comparatively clear, warranting stronger integrity credit than the sparse detail and simplified construction.","categories":[{"id":"fidelity","label":"Task & likeness","score":2.5},{"id":"structure","label":"Anatomy & construction","score":2.5},{"id":"space","label":"Space & placement","score":2.5},{"id":"light","label":"Light & reflections","score":1.5},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.0309268760914087,"y":0.29218296921389014,"scale":1.6160744021928897},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"tencent-hy4-preview-svg-raccoon-b4cedf42a4a26a4b","modelId":"tencent/hy4-preview","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-04","generatedAt":"2026-10-04T15:51:38.600Z","media":{"type":"image","src":"/benchmarks/media/020ef2a426ca79f8.png","alt":"Hy4 preview: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:51:38.600Z","quality":{"score":55.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"A lively cartoon communicates the animal pair traveling across water. Curving waves, rising spray and locally modeled gray forms strengthen motion and volume. Ambiguous seating, simplified joints and unclear control routing limit construction. Competing focal elements, decorative foam and inconsistent illumination leave the finish uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"tencent-hy4-preview-svg-pelican-b618cda0f539e818","modelId":"tencent/hy4-preview","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-04","generatedAt":"2026-10-04T15:51:38.600Z","media":{"type":"image","src":"/benchmarks/media/a734cf42252850d8.png","alt":"Hy4 preview: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:51:38.600Z","quality":{"score":60,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A recognizable, playful pelican cyclist with readable wheels, accessories and landscape layers. Form shading and ground shadows provide competent depth. Angular legs, uncertain pedal engagement and absent steering contact limit the riding construction; crowded overlaps and simple surfaces constrain refinement.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"tencent-hy4-preview-svg-beach-d030aa6f45fe19e9","modelId":"tencent/hy4-preview","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-04","generatedAt":"2026-10-04T15:51:38.600Z","media":{"type":"image","src":"/benchmarks/media/1c13230ea91eb39d.png","alt":"Hy4 preview: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:51:38.600Z","quality":{"score":60,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-22","rationale":"A cheerful beach illustration fulfills the visible subject requirements with coherent foreground construction and a harmonious sunset palette. Shallow depth, schematic reflections, ambiguous blue airborne forms and exposed geometry beneath the sunset limit its finish. Small accessories are readable, but materials and surface detail remain simplified.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-max-0902-svg-selfie-aea705fcb6511459","modelId":"qwen/qwen3.8-max-0902","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptHash":"aea705fcb6511459","date":"2026-10-04","generatedAt":"2026-10-04T15:52:26.509Z","media":{"type":"image","src":"/benchmarks/media/036a111b375cd086.png","alt":"Qwen3.8-Max-0902: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:52:26.509Z","quality":{"score":70.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A coherent ornamental illustration with balanced framing, appealing color and legible magical accessories. The disconnected raised wrist, incorrect raised-arm side and palm orientation, bear-like facial proportions and shallow material modeling limit execution. Clean decoration and compact spatial balance remain strengths.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummary":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"nvidia-nemotron-3-ultra-550b-a55b-svg-raccoon-b4cedf42a4a26a4b","modelId":"nvidia/nemotron-3-ultra-550b-a55b","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-04","generatedAt":"2026-10-04T16:03:55.222Z","media":{"type":"image","src":"/benchmarks/media/c67967de4854a13d.png","alt":"Nemotron 3 Ultra: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-04T16:03:55.222Z","quality":{"score":24.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The masked raccoon, dog and orange craft make the scene recognizable, with an appealing sunset palette. Overlapping bodies obscure the riding relationship, and the dog's connection to the controls remains unresolved. Weak water displacement, inconsistent shading and visible shape artifacts limit the finish. Ambiguous attachments do not establish detached limbs or inverted anatomy.","categories":[{"id":"fidelity","label":"Task & likeness","score":3.5},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":2},{"id":"detail","label":"Purposeful detail","score":3.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":3},{"id":"integrity","label":"Visual integrity","score":2.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"nvidia-nemotron-3-ultra-550b-a55b-svg-beach-d030aa6f45fe19e9","modelId":"nvidia/nemotron-3-ultra-550b-a55b","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-04","generatedAt":"2026-10-04T16:04:26.891Z","media":{"type":"image","src":"/benchmarks/media/80b904eb4c306820.png","alt":"Nemotron 3 Ultra: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:52:19.306Z","quality":{"score":39,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A recognizable beach illustration with appealing warm-cool contrast, readable layers and mostly clean shapes. Unresolved seat construction, schematic animals and broad geometric foliage give it an elementary finish. Weakly coordinated reflections and limited material differentiation further constrain depth and atmosphere.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":1.5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"nvidia-nemotron-3-ultra-550b-a55b-svg-selfie-5a721020202de5d8","modelId":"nvidia/nemotron-3-ultra-550b-a55b","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A hovering mechanical nautilus tests spiral construction, connected mechanical forms and contrasting metal and illuminated surfaces.","promptHash":"5a721020202de5d8","date":"2026-10-04","generatedAt":"2026-10-04T16:03:55.222Z","media":{"type":"image","src":"/benchmarks/media/eecf4fc8be2c4afc.png","alt":"Nemotron 3 Ultra: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-04T16:03:55.222Z","quality":{"score":36.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"An endearing mechanical creature is clearly framed, with expressive amber eyes and readable segmented arms. The shell resembles layered green plates more than a deep coiled spiral. Detached-looking construction, absent working contacts and patchy rendering leave the central assembly action substantially unresolved.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100,"visualScore":41.2,"technicalDeduction":5},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A hovering mechanical nautilus tests spiral construction, connected mechanical forms and contrasting metal and illuminated surfaces.","promptSummary":"A hovering mechanical nautilus tests spiral construction, connected mechanical forms and contrasting metal and illuminated surfaces.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"nvidia-nemotron-3-ultra-550b-a55b-svg-mona-lisa-e7b85d8cd393c26f","modelId":"nvidia/nemotron-3-ultra-550b-a55b","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-04","generatedAt":"2026-10-04T16:03:55.222Z","media":{"type":"image","src":"/benchmarks/media/ad03d1a8517a5e8d.png","alt":"Nemotron 3 Ultra: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-04T16:03:55.222Z","quality":{"score":0.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The centered silhouette provides basic organization, but multicolored mottling and geometric overlays leave the portrait unresolved. No readable face, crossed hands, clothing or landscape establishes Mona Lisa replication. Texture and transparency offer limited visual effects without coherent anatomy, lighting or material distinctions.","categories":[{"id":"fidelity","label":"Task & likeness","score":0.5},{"id":"structure","label":"Anatomy & construction","score":0},{"id":"space","label":"Space & placement","score":1},{"id":"light","label":"Light & reflections","score":0.5},{"id":"detail","label":"Purposeful detail","score":0.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":0},{"id":"integrity","label":"Visual integrity","score":0.5}],"maximum":100,"visualScore":5.5,"technicalDeduction":5},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04006006722664619,"y":0.22686995183525432,"scale":2.023846684922345},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"nvidia-nemotron-3-ultra-550b-a55b-svg-pelican-b618cda0f539e818","modelId":"nvidia/nemotron-3-ultra-550b-a55b","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-04","generatedAt":"2026-10-04T16:03:55.222Z","media":{"type":"image","src":"/benchmarks/media/36621ed6b45cf985.png","alt":"Nemotron 3 Ultra: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-04T16:03:55.222Z","quality":{"score":34.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A recognizable, playful pelican dominates a readable bicycle scene. Its bill and silhouette succeed, but feet below the pedals and confused drivetrain connections weaken the riding action. Local material modeling adds interest; overlapping contours, repetitive detail and inconsistent shadows leave the illustration basic and uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":3.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100,"visualScore":39.8,"technicalDeduction":5},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-max-0902-svg-pelican-b618cda0f539e818","modelId":"qwen/qwen3.8-max-0902","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-04","generatedAt":"2026-10-04T15:52:39.161Z","media":{"type":"image","src":"/benchmarks/media/61cd0866b3f3b4fa.png","alt":"Qwen3.8-Max-0902: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:52:39.161Z","quality":{"score":72,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"An appealing graphic illustration with a recognizable pelican, readable bicycle and clean coastal composition. Steering contact and one connected pedal support the action. Ambiguous neck-to-torso construction, an unclear opposing pedal connection, shallow space and limited material and lighting effects constrain its execution.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-max-0902-svg-mona-lisa-e7b85d8cd393c26f","modelId":"qwen/qwen3.8-max-0902","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-04","generatedAt":"2026-10-04T15:52:38.155Z","media":{"type":"image","src":"/benchmarks/media/9aaaaf7365c2fcd0.png","alt":"Qwen3.8-Max-0902: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:52:38.155Z","quality":{"score":45.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"A recognizable, cohesive Mona Lisa interpretation with readable landscape layers and a restrained expression. Continuous body construction and clean contours support readability, but broad hands, simplified drapery, frontal symmetry and limited facial volume keep the execution basic and uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":3.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.022517252340906853,"y":0.10471728461035514,"scale":1.616074402192891},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-max-0902-svg-raccoon-b4cedf42a4a26a4b","modelId":"qwen/qwen3.8-max-0902","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-04","generatedAt":"2026-10-04T15:53:02.625Z","media":{"type":"image","src":"/benchmarks/media/d3bd45ff3705d53b.png","alt":"Qwen3.8-Max-0902: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:53:02.625Z","quality":{"score":69.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-04","rationale":"The riding arrangement and steering action are readable, supported by coherent visible attachments and useful character details. Clean shapes, strong color separation and clustered foam make the scene appealing and active. Stiff anatomy, crowded controls, limited directional lighting and simple surface treatments constrain sophistication.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-max-0902-svg-beach-d030aa6f45fe19e9","modelId":"qwen/qwen3.8-max-0902","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-04","generatedAt":"2026-10-04T15:55:36.856Z","media":{"type":"image","src":"/benchmarks/media/390ed08830a87ce0.png","alt":"Qwen3.8-Max-0902: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:55:36.856Z","quality":{"score":77.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A strong stylized beach illustration with coherent sunset color, lively dolphin arcs and balanced activity. Clean contours, readable ground contact and an aligned reflection support the finish. Schematic anatomy, sparse small-subject features, flat materials and repetitive water highlights limit physical depth and refinement.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-1-sol-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-6.1-sol","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:32:34.032550+00:00","media":{"type":"image","src":"/benchmarks/media/30fac659c6069b30.png","alt":"GPT-6.1 Sol: Newton's Mirror SVG benchmark","thumbnail":"/assets/homepage/thumbnails/31c396394900f21f6d5ccf06.png"},"firstGeneratedAt":"2026-09-30T18:32:34.032550+00:00","quality":{"score":52.9,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Recognizable first sphere images and several central planar lights provide limited geometric agreement. Missing outer chains, truncated panel sequences and simplified interiors across the main, side and floor spheres leave most of the recursive reflection structure unsupported.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":24,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":15,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":35,"max":120},{"id":"materialsScale","label":"Materials and scale","score":10,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-astra-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-6-astra","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.559928+00:00","media":{"type":"image","src":"/benchmarks/media/3373cf328ebe8292.png","alt":"GPT-6 Astra: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.559928+00:00","quality":{"score":70.6,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-13","rationale":"Major placement and the visible contact sequence are convincing. Dense planar light repetition, recognizable spherical panel images and simplified nested sphere motifs provide useful optical structure. Missing outer chains, truncated distant sequences and incorrect bead-like belts still leave the recursive reflections substantially incomplete.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":25,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":15,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":70,"max":120},{"id":"materialsScale","label":"Materials and scale","score":10,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.00020070944668973065,"y":-0.040265402843601895,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-sol-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-6-sol","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.559928+00:00","media":{"type":"image","src":"/benchmarks/media/edc66568ae421770.png","alt":"GPT-6 Sol: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.559928+00:00","quality":{"score":30,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-04","rationale":"The centered sphere, separate airborne black form and color ordering support scene recognition. Rounded shading and curved rims provide readable volume and gloss. Shrinking side images, a floor counterpart and curved lamp highlights provide isolated reflection cues, but broad bands and sparse repetitions leave nested room imagery, outer chains and ordered emitter mapping unsupported. Vertical framing departs from the prescribed view, while paint contact, transport and solid-metal appearance remain weak or uncertain.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":19,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":12,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":8,"max":120},{"id":"materialsScale","label":"Materials and scale","score":6.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0014643626681248635,"y":0.011810690766113675,"scale":1.09417428370521},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-luna-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-6-luna","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.561439+00:00","media":{"type":"image","src":"/benchmarks/media/ca8a48278d44e0c7.png","alt":"GPT-6 Luna: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.561439+00:00","quality":{"score":26.6,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-04","rationale":"The schematic scene supports the paint contact order and compact paint scale. Its low, undersized main sphere and incorrectly placed wall images depart from the reference geometry. A few shrinking wall images provide limited optical organization, while absent lateral recursion and outer chains, broad sphere ribbons and isolated panel highlights leave most nested room and emitter mapping undeveloped.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":22,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":3.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.000985889871887997,"y":0.08174358579220933,"scale":0.8228346580560182},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-sol-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-5.6-sol","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.571024+00:00","media":{"type":"image","src":"/benchmarks/media/54859a7de489efb1.png","alt":"GPT-5.6 Sol: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.571024+00:00","quality":{"score":32.1,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-24","rationale":"The central sphere and paint ordering are supported, but a frontal, centrally converging chamber replaces the required corner perspective. The first wall and floor images have incorrect placement and scale relationships. Sparse reflection cues remain, while continuous outer chains, open lateral sightlines, dense planar light sequences and curved nested spherical imagery are largely missing.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":6,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":11,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.001935085002870673,"y":0.0015160295096287143,"scale":1.0304545339535167},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-terra-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-5.6-terra","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:01:48.572024+00:00","media":{"type":"image","src":"/benchmarks/media/fca25ce9923d89c6.png","alt":"GPT-5.6 Terra: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:01:48.572024+00:00","quality":{"score":27.3,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The real sphere and airborne black blob are substantially too low in the complete frame after accounting for aspect ratio.\nThe first wall and floor reflections have incompatible placement, relative size or silhouette.\nBoth lateral regions lack the references' receding lane-and-chain organization.\nThe few recognizable mirror images provide limited optical structure. The main, side and floor spheres replace nested mirror imagery with smooth bands, while sparse planar light patches and broad spherical highlights fail to establish coherent repeated emitter mapping.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":24,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":6,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":7,"max":120},{"id":"materialsScale","label":"Materials and scale","score":5.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0009083390722476231,"y":0.05607652773142518,"scale":1.015113064615719},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-luna-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-5.6-luna","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:05:39.372698+00:00","media":{"type":"image","src":"/benchmarks/media/b4ddf70c0f9c2ebe.png","alt":"GPT-5.6 Luna: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:05:39.372698+00:00","quality":{"score":21.9,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The room enclosure has coherent symmetric perspective, with the overhead panel, sphere and black aligned centrally. The sphere and airborne black occupy substantially lower positions than the reference framing, inconsistent with the prescribed camera aim. The diagonal crimson streak does not establish the required surface transport. Faint wall sphere images provide limited mirrored correspondence, but the shallow floor image, missing receding chains, absent nested sphere imagery and sparse emitter mapping leave most reflection structure unsupported.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":15,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":4,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":3.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":6.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.002546511302626851,"y":0.11038054687060252,"scale":1.127496851579375},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-5-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"openai/gpt-5.5","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:09:22.906932+00:00","media":{"type":"image","src":"/benchmarks/media/54ae766b17020fb5.png","alt":"GPT-5.5: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:09:22.906932+00:00","quality":{"score":22.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The paint ordering and separate black bead support part of the physical state, but adhesion and rotational transport remain only partly established. The sweeping cyan arc and long crimson lobes do not clearly communicate the limited time since their respective impacts. The main sphere is too large and too low, and the black bead is below its required projected position. Displaced isolated sphere images, a flattened floor echo and missing recursive, nested and emitter reflections provide little coherent optical structure. The colored paint is excessively bulky for the specified quantities.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":15,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":6,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":6,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.005587335059814878,"y":0.09653689688396669,"scale":1.3498588075760019},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-newtons-mirror-5a00e74f7bfe6e8e-from-20260929","modelId":"space-bunny-alpha","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:45:44.580Z","edition":{"id":"from-20260929","label":"29 Sep onward"},"media":{"type":"image","src":"/benchmarks/media/4aa025b5d387fa5a.png","alt":"Space Bunny Alpha: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:45:44.580Z","quality":{"score":24.9,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Visible panel shapes, curved sphere highlights, rounded side images and a lower continuation beneath the large left image establish recognizable local mirror structure. Their uneven arrangement, missing receding chains and absent nested imagery remain substantial shortcomings. The main sphere's framing is close, but black lacks projected separation and cyan does not establish recent apex contact.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":11,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":9,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":15.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1","visualScore":29.9,"technicalDeduction":5},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"owl-alpha-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"owl-alpha","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:45:46.458Z","media":{"type":"image","src":"/benchmarks/media/0be455e6e49e58f0.png","alt":"Owl Alpha: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:45:46.458Z","quality":{"score":23.1,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-13","rationale":"Cyan lies on the sphere's facing middle rather than showing its recent interaction near the apex. Separate airborne black and elongated crimson support part of the timing sequence. Small side images and an overlapping lower image provide limited reflection ordering, but recursive room structure, nested imagery and emitter mappings are largely absent. The main sphere is smaller and higher in the frame than prescribed, and the surfaces have a soft, matte appearance.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":9.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":6.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":12.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":5.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.004428239873428717,"y":-0.03883709581949477,"scale":0.8228346580560183},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"ox-alpha-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"ox-alpha","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T20:17:11.602Z","media":{"type":"image","src":"/benchmarks/media/1e3201bbe784c260.png","alt":"Ox Alpha: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T20:17:11.602Z","quality":{"score":27.6,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The setup is recognizable, but the paint interaction remains poorly established and off-center. The first wall and floor images provide coherent placement, scale, silhouette and occlusion cues, with faint lower echoes. Missing receding chains, nested imagery and ordered emitter mappings leave the reflection system heavily simplified. Dull shading does not convincingly depict polished steel or mirrors.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":14,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":9,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":21.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"union-alpha-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"union-alpha","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:45:43.779Z","media":{"type":"image","src":"/benchmarks/media/f744f4359b04d097.png","alt":"Union Alpha: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:45:43.779Z","quality":{"score":52,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Recognizable near-field sphere placement and bilateral convergence provide limited optical structure. Broad gray regions interrupt lateral sightlines and truncate side-floor images. Missing outer chains, absent nested sphere imagery and incomplete planar and curved emitter mappings dominate the reflection system.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":24,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":13.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":34.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":10,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-pro-0813-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"deepseek/deepseek-v4-pro-0813","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:45:44.293Z","media":{"type":"image","src":"/benchmarks/media/186b99118f0f753e.png","alt":"DeepSeek-V4-Pro-0813: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:45:44.293Z","quality":{"score":36.9,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The central sphere and gross paint state are coherent. Reduced circular side images and paired overhead quadrilaterals offer limited, spatially organized reflection cues, but the floor-sphere image, receding mirror-room structure, outer chains, nested imagery and curved emitter mappings are largely missing or invalid.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":22,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":15,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":11,"max":120},{"id":"materialsScale","label":"Materials and scale","score":8,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"deepseek/deepseek-v4.1-flash","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:13:04.603Z","media":{"type":"image","src":"/benchmarks/media/9188ef18ade85b52.png","alt":"DeepSeek-V4.1-Flash: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:13:04.603Z","quality":{"score":46.2,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The first wall and floor images provide limited reflection geometry, but the room's mirror recurrence is substantially incomplete, particularly the outer-edge chains and lateral sightlines. The main sphere and its large reflected images lack ordered, curved room and virtual-sphere imagery. Repeated emitter mapping is missing from both planar mirrors and spherical surfaces.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":13.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":30,"max":120},{"id":"materialsScale","label":"Materials and scale","score":8,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-flash-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"qwen/qwen3.8-flash","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-01","generatedAt":"2026-10-01T21:35:33.819Z","media":{"type":"image","src":"/benchmarks/media/dc203e194a9f1538.png","alt":"Qwen3.8-Flash: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-01T21:35:33.819Z","quality":{"score":22.4,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The full-frame illustration supports a central steel sphere, an airborne black blob and the relative ordering of cyan and crimson. Its faint floor return and symmetric repetitions provide sparse mirror structure; outer chains, large reflected images, nested imagery and emitter mapping remain deficient. Vertical framing is substantially displaced, and paint contact histories remain unresolved.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":19,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":3.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":9.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":4,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0006647237405337936,"y":0.11203485765038096,"scale":0.9704455335485082},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-3-8-flash-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"google/gemini-3.8-flash","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:04:28.962Z","media":{"type":"image","src":"/benchmarks/media/5b744d02760745e9.png","alt":"Gemini 3.8 Flash: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:04:28.962Z","quality":{"score":30.3,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-04","rationale":"Crimson is carried well down the camera-facing surface, supporting its earlier interaction. The depiction remains highly simplified, with imperfect material rendering. Isolated wall and floor images and a few highlights provide scant reflective evidence; displaced first reflections, missing receding chains and absent nested sphere and emitter imagery leave the room's optical structure largely unsupported.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":22.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":11,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":4.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"x-ai-grok-4-7-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"x-ai/grok-4.7","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:24:29.919Z","media":{"type":"image","src":"/benchmarks/media/d085a469a3248a25.png","alt":"Grok 4.7: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:24:29.919Z","quality":{"score":41,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-13","rationale":"Some sphere reflections are correctly placed, with useful size progression and a floor image. Their overlapping side continuations and inverted floor paint order provide additional local optical cues. The gross paint sequence is recognizable, although crimson is a little too low. Incorrect lamp projections and absent nested and receding imagery still leave the mirror system largely incomplete.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":12.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":26,"max":120},{"id":"materialsScale","label":"Materials and scale","score":6.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"minimax/minimax-m3","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T18:45:43.863Z","media":{"type":"image","src":"/benchmarks/media/af48c1a3e8b542c9.png","alt":"MiniMax M3: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T18:45:43.863Z","quality":{"score":29.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The centered sphere, aligned left wall image, smaller lower-left continuation and partially occluded floor image provide recognizable local mirror coherence. Cyan remains airborne, the main sphere resembles white marble, and missing right-side imagery, recursive optics and panel mappings leave the scene substantially incomplete.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":7,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":14.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":22,"max":120},{"id":"materialsScale","label":"Materials and scale","score":5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-sonnet-5-5-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"anthropic/claude-sonnet-5.5","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:15:08.866Z","media":{"type":"image","src":"/benchmarks/media/a673cfbab2c0984f.png","alt":"Claude Sonnet 5.5: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:15:08.866Z","quality":{"score":52.9,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The near-field silhouettes, broad perspective and paint contact order are recognizable. Missing outer chains, truncated repetitions and absent ordered imagery within the main and reflected spheres substantially limit the reflection system. Sparse spherical highlights leave much of the curved room and emitter mapping unsupported, despite some plausible planar light sequences.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":24,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":15,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":35,"max":120},{"id":"materialsScale","label":"Materials and scale","score":10,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.000920245398773006,"y":0.007361963190184048,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-flash-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"xiaomi/mimo-v2.6-flash","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:56:10.214Z","media":{"type":"image","src":"/benchmarks/media/b057dbac96e75d1c.png","alt":"MiMo-V2.6-Flash: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:56:10.214Z","quality":{"score":12,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The faint lower sphere image provides limited reflection evidence through reversed paint-color ordering. The wider reflection system and nested room imagery are absent. The colored beads do not resolve the required impact and transport history, while displaced paint positions and weak material rendering further limit the result.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":4.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":6.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":2.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.003148748433911427,"y":-0.015221781500072707,"scale":0.8607079764250576},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-27b-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"qwen/qwen3.8-27B","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-01","generatedAt":"2026-10-01T23:34:46.728Z","media":{"type":"image","src":"/benchmarks/media/12bf4e126286660a.png","alt":"Qwen3.8-27B: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-01T23:34:46.728Z","quality":{"score":40.2,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The room fills the complete 16:9 frame without visible borders or labels. Central placement fits the wider frame, and the paint colors follow the expected vertical order. Bilateral side images, diminishing echoes and an overlapping floor image provide limited mirror depth. Broad sphere surfaces lack ordered nested room imagery, while planar panel sequences and outer chains remain missing or unresolved. Soft highlights do not establish emitter mapping, while paint attachment and transport history are only partly supported.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":13,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":24,"max":120},{"id":"materialsScale","label":"Materials and scale","score":6.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-fable-5-1-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"anthropic/claude-fable-5.1","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:15:08.866Z","media":{"type":"image","src":"/benchmarks/media/2cd1205052059d72.png","alt":"Claude Fable 5.1: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:15:08.866Z","quality":{"score":54.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The broad paint sequence, bilateral near-field sphere arrangement, overlaps and diminishing panel repetitions are recognizable; crimson is slightly too low. Small paint scale and basic shading provide material cues, although the steel response remains soft. Inaccurate light projections, missing outer chains and incomplete nested room imagery remain the principal limitations.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":24,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":14,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":48,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-opus-5-5-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"anthropic/claude-opus-5.5","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:40:08.798Z","media":{"type":"image","src":"/benchmarks/media/9fe8b1cebf732f3c.png","alt":"Claude Opus 5.5: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:40:08.798Z","quality":{"score":61.1,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Open lateral sightlines, first reflected silhouettes and near-field planar panel placement are convincing. Missing outer chains, truncated repetitions and absent nested spherical room and emitter imagery remain the main weaknesses.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":24,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":15,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":55,"max":120},{"id":"materialsScale","label":"Materials and scale","score":9,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"z-ai-glm-5-3-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"z-ai/glm-5.3","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T20:35:51.447Z","media":{"type":"image","src":"/benchmarks/media/ba2eb52799a7cab2.png","alt":"GLM-5.3: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T20:35:51.447Z","quality":{"score":25.4,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The sphere has readable rounded volume, and the compact black blob remains distinctly above it, supporting pre-impact separation. The extensive red and cyan coatings do not establish the required interaction history or paint scale. The sphere and black are substantially displaced downward. A centered upper panel, paired side patches and a vertically aligned partial floor image provide limited reflection cues, but ordered mirror-room repetitions and curved spherical imagery are largely missing.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":9,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":9,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":10,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0004636176787719458,"y":0.11368853040315142,"scale":1.1972173631218093},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"moonshotai/kimi-k3","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-02","generatedAt":"2026-10-02T01:50:03.361Z","media":{"type":"image","src":"/benchmarks/media/31b9402377af76cc.png","alt":"Kimi K3: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-02T01:50:03.361Z","quality":{"score":35.3,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-13","rationale":"The room fills the complete portrait frame without visible borders or labels. The main sphere and airborne black are convincingly placed within the portrait frame. Paint ordering is supported, but its contact and transport history remains weakly expressed. Symmetric wall ghosts, inverted floor imagery and panel echoes establish a basic mirror system. Diffuse gradients dominate, with little ordered nested imagery or repeated emitter mapping on either planar or spherical surfaces. The wall images sit too low, and the floor image is unusually large and low.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":18,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":8,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":24,"max":120},{"id":"materialsScale","label":"Materials and scale","score":6.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.003164327679253399,"y":0.003029356356441643,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-1-3-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"meta/muse-spark-1.3","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-09-30","generatedAt":"2026-09-30T19:21:49.013Z","media":{"type":"image","src":"/benchmarks/media/ea732f9bd0a83eb5.png","alt":"Muse Spark 1.3: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-09-30T19:21:49.013Z","quality":{"score":16.5,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The paint state is incorrectly and poorly drawn, despite a recognizable sphere-and-paint setup. The oversized sphere, oversized compact paint bodies, and blurred wall and floor images further limit the result. Isolated reflection cues provide very little support for the required repeated virtual rooms, nested spherical imagery or emitter mapping.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":6,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":8.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":5.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":4,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.0004088138314292187,"y":0.009864468762627051,"scale":1.3909681284637787},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"pixel-canary-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"pixel-canary","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-01","media":{"type":"image","src":"/benchmarks/media/0813427500b4cc0c.png","alt":"Pixel Canary: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-01","quality":{"score":32.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The principal sphere placement and paint contact sequence are supported. Isolated sphere and panel images provide limited optical structure, but missing lateral chains and side-floor continuations prevent a coherent repeating mirror room. The main and reflected spheres lack ordered curved nested imagery, while sparse planar emitter images and incomplete spherical emitter mapping leave most of the reflection system unsupported.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":20,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":13,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":10,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"fledge-alpha-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"fledge-alpha","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-02","generatedAt":"2026-10-02T00:40:36.900Z","media":{"type":"image","src":"/benchmarks/media/82402e8b98a6bec0.png","alt":"Fledge Alpha: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-02T00:40:36.900Z","quality":{"score":16.4,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-02","rationale":"The candidate communicates a round metallic sphere, three paint colors and a separate airborne black blob. Its hovering support relationship, displaced framing and unresolved paint interaction limit physical credit. Sparse planar light ghosts and a soft sphere highlight provide little evidence for the required wall/floor repetitions and curved nested mirror imagery.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":9,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":7.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.004242824783070308,"y":0.06373624699003459,"scale":1.161834242728283},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"nvidia-nemotron-3-ultra-550b-a55b-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"nvidia/nemotron-3-ultra-550b-a55b","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-04","generatedAt":"2026-10-04T16:04:42.145Z","media":{"type":"image","src":"/benchmarks/media/dcef7bf6a219c349.png","alt":"Nemotron 3 Ultra: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-04T16:04:42.145Z","quality":{"score":9.9,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The candidate establishes a separate airborne black blob and partial crimson overlap, but leaves the paint interaction history and settled support unresolved. Its main silhouette, black placement and lamp projection depart substantially from the required camera geometry. A few faint reflection hints provide little support for the extensive mirror system, while the compact paint volumes are visibly disproportionate.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":2,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":3.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":4.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.002342505895513482,"y":0.024015578950800687,"scale":0.5406408953093166},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"tencent-hy4-preview-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"tencent/hy4-preview","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-04","generatedAt":"2026-10-04T15:51:38.600Z","media":{"type":"image","src":"/benchmarks/media/44195a11ccf2edde.png","alt":"Hy4 preview: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:51:38.600Z","quality":{"score":41.4,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-13","rationale":"The central sphere, compact airborne black bead and broad paint ordering are supported. Contact and transport history remain partly unresolved. The optical system provides only isolated first images and light patches, with little evidence of the extensive ordered reflection structure. Rounded metal and paint gloss are more successful than the reflected contents.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":18,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":11.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":29.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-max-0902-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"qwen/qwen3.8-max-0902","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-04","generatedAt":"2026-10-04T15:55:43.156Z","media":{"type":"image","src":"/benchmarks/media/48f3e3b0214c154f.png","alt":"Qwen3.8-Max-0902: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-04T15:55:43.156Z","quality":{"score":44.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The central sphere and differentiated cyan/crimson interaction are supported. The tiny highlighted object above the sphere provides partial airborne-black evidence. The few isolated sphere and panel images preserve limited optical relationships, but most of the repeating mirror-room structure is missing. Core score: 64.5/170; including the floor-damage reward: 64.5/180.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":12,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":29,"max":120},{"id":"materialsScale","label":"Materials and scale","score":8,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"aleph-alpha-kolibri-1-svg-selfie-e5bcbc64de37f34f","modelId":"Aleph-Alpha/Kolibri-1","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A jewel-coloured hummingbird hovering in flight tests avian anatomy, delicate structure and iridescent surface detail.","promptHash":"e5bcbc64de37f34f","date":"2026-10-05","generatedAt":"2026-10-05T16:55:04.716Z","media":{"type":"image","src":"/benchmarks/media/fba0e930cae81873.png","alt":"Kolibri-1: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-05T16:55:04.716Z","quality":{"score":22.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A restrained ivory, teal and green palette gives the image simple graphic coherence. Execution remains rudimentary: detached oval wings, ambiguous central anatomy and unclear feeding contact prevent a convincing hummingbird scene. The requested feather structure, botanical detail, lighting and enclosing arch are largely unresolved.","categories":[{"id":"fidelity","label":"Task & likeness","score":2.5},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":3},{"id":"light","label":"Light & reflections","score":2},{"id":"detail","label":"Purposeful detail","score":1.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jewel-coloured hummingbird hovering in flight tests avian anatomy, delicate structure and iridescent surface detail.","promptSummary":"A jewel-coloured hummingbird hovering in flight tests avian anatomy, delicate structure and iridescent surface detail.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"aleph-alpha-kolibri-1-svg-mona-lisa-e7b85d8cd393c26f","modelId":"Aleph-Alpha/Kolibri-1","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-05","generatedAt":"2026-10-05T16:56:45.734Z","media":{"type":"image","src":"/benchmarks/media/3836dfe3fce4c0e2.png","alt":"Kolibri-1: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-05T16:56:45.734Z","quality":{"score":3.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"The candidate offers a clean, highly simplified frontal portrait with warm skin and dark hair. Its oversized oval face, schematic features and unresolved body provide little Mona Lisa likeness. Missing crossed hands, clothing detail and landscape remove essential parts of the original composition; smooth gradients supply only limited volume.","categories":[{"id":"fidelity","label":"Task & likeness","score":0.5},{"id":"structure","label":"Anatomy & construction","score":0.5},{"id":"space","label":"Space & placement","score":0},{"id":"light","label":"Light & reflections","score":0},{"id":"detail","label":"Purposeful detail","score":0},{"id":"artistry","label":"Artistry & character","score":0.5},{"id":"effects","label":"Materials & effects","score":0.5},{"id":"integrity","label":"Visual integrity","score":0.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.09050977870911492,"y":0.4372402358805609,"scale":2.2144409968040724},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"aleph-alpha-kolibri-1-svg-beach-d030aa6f45fe19e9","modelId":"Aleph-Alpha/Kolibri-1","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-05","generatedAt":"2026-10-05T16:58:33.076Z","media":{"type":"image","src":"/benchmarks/media/9f0b429aab7cc30f.png","alt":"Kolibri-1: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-05T16:58:33.076Z","quality":{"score":12.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A recognizable sunset beach has an appealing purple-orange gradient and clean geometric edges. Birds and animals remain unresolved, terrain obscures the relationship between shore and water, and reflections are absent. Sparse detail, flat materials and limited activity keep the drawing elementary despite its tidy rendering.","categories":[{"id":"fidelity","label":"Task & likeness","score":2.5},{"id":"structure","label":"Anatomy & construction","score":0.5},{"id":"space","label":"Space & placement","score":3},{"id":"light","label":"Light & reflections","score":1},{"id":"detail","label":"Purposeful detail","score":0.5},{"id":"artistry","label":"Artistry & character","score":1},{"id":"effects","label":"Materials & effects","score":0.5},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"aleph-alpha-kolibri-1-svg-raccoon-b4cedf42a4a26a4b","modelId":"Aleph-Alpha/Kolibri-1","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-05","generatedAt":"2026-10-05T16:58:30.883Z","media":{"type":"image","src":"/benchmarks/media/e005d9c8a607783d.png","alt":"Kolibri-1: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-05T16:58:30.883Z","quality":{"score":10,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Clean shapes and a readable coral platform provide basic graphic clarity, but the animal construction remains fragmented. A recognizable raccoon riding the dog's back, connected steering and convincing water contact are not demonstrated. Sparse detail, flat subjects and absent movement effects support retaining the baseline category assessments.","categories":[{"id":"fidelity","label":"Task & likeness","score":1.5},{"id":"structure","label":"Anatomy & construction","score":1},{"id":"space","label":"Space & placement","score":1},{"id":"light","label":"Light & reflections","score":0},{"id":"detail","label":"Purposeful detail","score":0.5},{"id":"artistry","label":"Artistry & character","score":0.5},{"id":"effects","label":"Materials & effects","score":0.5},{"id":"integrity","label":"Visual integrity","score":3.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"aleph-alpha-kolibri-1-svg-pelican-b618cda0f539e818","modelId":"Aleph-Alpha/Kolibri-1","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-05","generatedAt":"2026-10-05T16:58:30.143Z","media":{"type":"image","src":"/benchmarks/media/d7d2fd372659440e.png","alt":"Kolibri-1: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-05T16:58:30.143Z","quality":{"score":17.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Clean shapes and aligned wheels establish a readable schematic scene. Pelican likeness, riding support and bicycle mechanics remain substantially unresolved. Sparse detail and flat surfaces limit expression and volume. Strong edge and fill integrity is a separate success and does not repair the construction.","categories":[{"id":"fidelity","label":"Task & likeness","score":1.5},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":1.5},{"id":"light","label":"Light & reflections","score":3},{"id":"detail","label":"Purposeful detail","score":1.5},{"id":"artistry","label":"Artistry & character","score":1},{"id":"effects","label":"Materials & effects","score":1},{"id":"integrity","label":"Visual integrity","score":3.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"aleph-alpha-kolibri-1-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"Aleph-Alpha/Kolibri-1","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-05","generatedAt":"2026-10-05T16:58:32.268Z","media":{"type":"image","src":"/benchmarks/media/fe37f45d4946fca9.png","alt":"Kolibri-1: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-05T16:58:32.268Z","quality":{"score":10.4,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"A segmented gray perimeter surrounds a black field containing one large, low-positioned gray sphere and two small colored circles. Rounded volume and restrained paint scale earn partial material credit. The image does not demonstrate the required optical structure, while paint history, airborne black and settled support remain largely unresolved.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":9.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":1.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":0,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.0016133794833723147,"y":0.27136906375019443,"scale":1.5683121854901683},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-35b-a3b-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"qwen/qwen3.6-35b-a3b","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-05","generatedAt":"2026-10-05T17:06:30.811Z","media":{"type":"image","src":"/benchmarks/media/22825c786afd03df.png","alt":"Qwen3.6-35B-A3B: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:06:30.811Z","quality":{"score":9.1,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The artwork demonstrates rounded volume, recognizable paint colors and a compact black feature at an appropriate screen height. The enlarged main-sphere framing, unsupported crimson and cyan interaction histories, oversized compact paint balls and largely missing mirror structure remain substantial weaknesses. The core assessment is 20/170; with no demonstrated floor damage, the combined assessment remains 20/180.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":2.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":4,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":3.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":2.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.0005147567549575254,"y":-0.022890273130660357,"scale":1.9059870292719217},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-35b-a3b-svg-raccoon-b4cedf42a4a26a4b","modelId":"qwen/qwen3.6-35b-a3b","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-05","generatedAt":"2026-10-05T17:06:30.811Z","media":{"type":"image","src":"/benchmarks/media/1e09fb1a8a360a9f.png","alt":"Qwen3.6-35B-A3B: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:06:30.811Z","quality":{"score":43.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"Clean shapes, cheerful faces and bright colors make the riding arrangement readable. The dog does not visibly operate the control, while awkward limbs, shallow hull placement and stiff poses limit credibility. Gradients and layered foam add modest volume and motion. The obscured tail attachment does not establish detachment.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-35b-a3b-svg-selfie-aea705fcb6511459","modelId":"qwen/qwen3.6-35b-a3b","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptHash":"aea705fcb6511459","date":"2026-10-05","generatedAt":"2026-10-05T17:06:30.810Z","media":{"type":"image","src":"/benchmarks/media/859acddc2b0cc233.png","alt":"Qwen3.6-35B-A3B: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:06:30.810Z","quality":{"score":63.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"An appealing fantasy illustration with coordinated ornament, readable suspension and luminous turquoise accents. Plausible limb attachments and soft shading support competent execution. Bear-like proportions, unclear paw gestures, insufficient white robe fabric, crowded chest details and shallow material modeling limit its finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummary":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-35b-a3b-svg-mona-lisa-e7b85d8cd393c26f","modelId":"qwen/qwen3.6-35b-a3b","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-05","generatedAt":"2026-10-05T17:06:30.810Z","media":{"type":"image","src":"/benchmarks/media/74277f840c88984d.png","alt":"Qwen3.6-35B-A3B: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:06:30.810Z","quality":{"score":13.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A recognizable but basic geometric interpretation. Muted colors and soft shading provide cohesion, while weak likeness, unresolved hands, rigid anatomy and shallow landscape construction substantially limit the replication.","categories":[{"id":"fidelity","label":"Task & likeness","score":2},{"id":"structure","label":"Anatomy & construction","score":1},{"id":"space","label":"Space & placement","score":2},{"id":"light","label":"Light & reflections","score":0.5},{"id":"detail","label":"Purposeful detail","score":1.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":1.5},{"id":"integrity","label":"Visual integrity","score":2}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04042298801737344,"y":0.10067068285370238,"scale":1.3099644507332462},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-35b-a3b-svg-pelican-b618cda0f539e818","modelId":"qwen/qwen3.6-35b-a3b","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-05","generatedAt":"2026-10-05T17:06:30.811Z","media":{"type":"image","src":"/benchmarks/media/84345358683296ed.png","alt":"Qwen3.6-35B-A3B: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:06:30.811Z","quality":{"score":53.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cheerful, readable illustration with a distinctive pelican silhouette, clean contours and an orderly landscape. Soft glow and restrained shading provide modest atmosphere. Construction is weaker: the handlebar is disconnected, wings do not reach the controls, and overlapping legs leave pedaling ambiguous. Simple forms and mostly flat materials limit depth and drawing sophistication.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-35b-a3b-svg-beach-d030aa6f45fe19e9","modelId":"qwen/qwen3.6-35b-a3b","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-05","generatedAt":"2026-10-05T17:06:30.811Z","media":{"type":"image","src":"/benchmarks/media/4eb0e53861aba57b.png","alt":"Qwen3.6-35B-A3B: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:06:30.811Z","quality":{"score":50.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A recognizable beach sunset with balanced framing and appealing color transitions. Birds, boats and small animals fulfill the main scene requirements. Elementary anatomy, faint details, weak grounding and rigid reflection columns limit depth and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-5-9b-svg-mona-lisa-e7b85d8cd393c26f","modelId":"qwen/qwen3.5-9b","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-05","generatedAt":"2026-10-05T17:11:33.563Z","media":{"type":"image","src":"/benchmarks/media/4cd5c45c42718e69.png","alt":"Qwen3.5-9B: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:11:33.563Z","quality":{"score":1.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"A centered dark oval with sparse facial marks forms a simple graphic face, but scarcely replicates the Mona Lisa. The portrait's body, crossed hands, clothing and landscape are missing, while angular facial construction and flat tones lose its likeness, volume and subtle expression.","categories":[{"id":"fidelity","label":"Task & likeness","score":0},{"id":"structure","label":"Anatomy & construction","score":0},{"id":"space","label":"Space & placement","score":0},{"id":"light","label":"Light & reflections","score":0},{"id":"detail","label":"Purposeful detail","score":0},{"id":"artistry","label":"Artistry & character","score":0.5},{"id":"effects","label":"Materials & effects","score":0},{"id":"integrity","label":"Visual integrity","score":0.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.08589860705157747,"y":0.7274494549710524,"scale":2.900838984059627},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-5-9b-svg-pelican-b618cda0f539e818","modelId":"qwen/qwen3.5-9b","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-05","generatedAt":"2026-10-05T17:11:33.563Z","media":{"type":"image","src":"/benchmarks/media/cbe1dd40152b9f75.png","alt":"Qwen3.5-9B: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:11:33.563Z","quality":{"score":23,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The oversized bill and orderly spoked wheels make the subject partly recognizable. Clean contours and strong color contrast provide graphic clarity, but unresolved appendages, steering, drivetrain and support dominate the construction. Sparse shading and detail further limit the illustration.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":3},{"id":"light","label":"Light & reflections","score":1.5},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":1.5},{"id":"integrity","label":"Visual integrity","score":2.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-5-9b-svg-raccoon-b4cedf42a4a26a4b","modelId":"qwen/qwen3.5-9b","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-05","generatedAt":"2026-10-05T17:11:33.564Z","media":{"type":"image","src":"/benchmarks/media/7a2b0d44b97ba3a2.png","alt":"Qwen3.5-9B: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:11:33.564Z","quality":{"score":14.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Bright colors, clean individual shapes and basic animal cues make the scene partly recognizable. Back seating, body continuity, steering and jet ski construction remain unresolved. Background gradients and wave rhythm provide some finish, but substantial anatomical, spatial and interaction weaknesses dominate.","categories":[{"id":"fidelity","label":"Task & likeness","score":2.5},{"id":"structure","label":"Anatomy & construction","score":1},{"id":"space","label":"Space & placement","score":2},{"id":"light","label":"Light & reflections","score":1},{"id":"detail","label":"Purposeful detail","score":1.5},{"id":"artistry","label":"Artistry & character","score":1},{"id":"effects","label":"Materials & effects","score":1},{"id":"integrity","label":"Visual integrity","score":2.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-5-9b-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"qwen/qwen3.5-9b","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-05","generatedAt":"2026-10-05T17:11:33.564Z","media":{"type":"image","src":"/benchmarks/media/92cfd01166c7b9ce.png","alt":"Qwen3.5-9B: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:11:33.564Z","quality":{"score":9.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Horizontal centering and a separate airborne black blob are visible. Cyan's recent contact and crimson's forward transport are insufficiently demonstrated, while sphere support remains obscured. The enlarged, lowered sphere substantially changes the composition. The wall and floor sphere images, recursive room structure and curved emitter mappings are almost entirely missing.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":3,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":4.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":2,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.00047743508003652256,"y":0.1274604788683303,"scale":1.39096812846378},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-5-9b-svg-selfie-aea705fcb6511459","modelId":"qwen/qwen3.5-9b","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptHash":"aea705fcb6511459","date":"2026-10-05","generatedAt":"2026-10-05T17:11:33.563Z","media":{"type":"image","src":"/benchmarks/media/0d9108660a349fdd.png","alt":"Qwen3.5-9B: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:11:33.563Z","quality":{"score":25.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"Glowing teal spheres and a harmonious palette provide visible strengths, but the central character remains seriously unresolved. A large gap separates head and torso, limbs cannot be traced, and scattered garment shapes fail to establish the ornate costume. The foreground mound replaces the tall enclosing arch. Clean primitive edges do not repair these construction failures.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":3},{"id":"light","label":"Light & reflections","score":3.5},{"id":"detail","label":"Purposeful detail","score":2},{"id":"artistry","label":"Artistry & character","score":2},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummary":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-5-9b-svg-beach-d030aa6f45fe19e9","modelId":"qwen/qwen3.5-9b","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-05","generatedAt":"2026-10-05T17:14:00.073Z","media":{"type":"image","src":"/benchmarks/media/f6ca1994f9d37e65.png","alt":"Qwen3.5-9B: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-05T17:14:00.073Z","quality":{"score":12.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The beach is recognizable, and sweeping shoreline curves provide basic rhythm. Wildlife and playful activity remain poorly resolved, with ambiguous body construction and object contact. The sun glows without corresponding reflections or surface lighting. Fragmented subjects and an unintegrated rectangular panel substantially weaken composition and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":1.5},{"id":"structure","label":"Anatomy & construction","score":0.5},{"id":"space","label":"Space & placement","score":1.5},{"id":"light","label":"Light & reflections","score":1},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":0.5},{"id":"integrity","label":"Visual integrity","score":1.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"inclusionai-ling-3-1-flash-svg-raccoon-b4cedf42a4a26a4b","modelId":"inclusionai/ling-3.1-flash","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-05","generatedAt":"2026-10-05T19:33:14.039Z","media":{"type":"image","src":"/benchmarks/media/277f2e3f3b0f2d14.png","alt":"Ling 3.1 Flash: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-05T19:33:14.039Z","quality":{"score":63.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"An appealing sunset cartoon with recognizable animals, expressive faces and coherent landscape layers. The sky and warm water streaks provide its strongest rendering. Ambiguous seated support, stiff forelegs, a shallow hull and an unclear steering connection limit construction. Clean shapes and readable accessories add finish, while restrained spray and flat surfaces limit material and motion effects.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"inclusionai-ling-3-1-flash-svg-beach-d030aa6f45fe19e9","modelId":"inclusionai/ling-3.1-flash","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-05","generatedAt":"2026-10-05T20:53:26.528Z","media":{"type":"image","src":"/benchmarks/media/83cf365e9fbe3649.png","alt":"Ling 3.1 Flash: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-05T20:53:26.528Z","quality":{"score":65.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A coherent sunset illustration with harmonious color, clean silhouettes and readable beach details. Its calm composition works well, while playful activity is modest. Simplified animal anatomy, shallow shoreline layering, repetitive foliage and a geometric reflection limit drawing sophistication and naturalism.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"inclusionai-ling-3-1-flash-svg-mona-lisa-e7b85d8cd393c26f","modelId":"inclusionai/ling-3.1-flash","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-05","generatedAt":"2026-10-05T21:10:14.811Z","media":{"type":"image","src":"/benchmarks/media/dcd436d5c5459f6d.png","alt":"Ling 3.1 Flash: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-05T21:10:14.811Z","quality":{"score":24,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"A basic, uneven Mona Lisa interpretation with a clear facial focal point, articulated eyes and modest tonal volume. Weak likeness, oversized head proportions, unclear shoulders, curled bare arms and indistinct hands limit replication. The compressed composition and rectangular torso weaken an otherwise coherent earthy palette.","categories":[{"id":"fidelity","label":"Task & likeness","score":2},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":3.5},{"id":"detail","label":"Purposeful detail","score":3.5},{"id":"artistry","label":"Artistry & character","score":2},{"id":"effects","label":"Materials & effects","score":1.5},{"id":"integrity","label":"Visual integrity","score":2}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.035198576037379026,"y":0.25665628360588866,"scale":1},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"inclusionai-ling-3-1-flash-svg-pelican-b618cda0f539e818","modelId":"inclusionai/ling-3.1-flash","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-05","generatedAt":"2026-10-05T21:15:57.106Z","media":{"type":"image","src":"/benchmarks/media/40044dd5fe7e053e.png","alt":"Ling 3.1 Flash: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-05T21:15:57.106Z","quality":{"score":64.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"A clean, appealing pelican illustration with a readable bicycle and coherent landscape. Bill shading, feather layers and faint wheel shadows provide modest depth. Uncertain pedal contact, simplified leg construction and limited material rendering keep the result within competent illustration quality.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"inclusionai-ling-3-1-flash-svg-selfie-cd427d065123fd30","modelId":"inclusionai/ling-3.1-flash","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An ant tending signals in a miniature workshop tests insect anatomy, purposeful gestures and the placement of small objects in a lit environment.","promptHash":"cd427d065123fd30","date":"2026-10-05","generatedAt":"2026-10-05T22:14:27.004Z","media":{"type":"image","src":"/benchmarks/media/927c3affdfff8e3b.png","alt":"Ling 3.1 Flash: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-05T22:14:27.004Z","quality":{"score":73.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"An appealing, coherent storybook ant with a clear expression, complete silhouette and readable signal mechanism. Clean curves and warm accents support the composition, while simplified joints, schematic hardware, limited material differentiation and modest light interaction constrain its finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An ant tending signals in a miniature workshop tests insect anatomy, purposeful gestures and the placement of small objects in a lit environment.","promptSummary":"An ant tending signals in a miniature workshop tests insect anatomy, purposeful gestures and the placement of small objects in a lit environment.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"inclusionai-ling-3-1-flash-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"inclusionai/ling-3.1-flash","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-05","generatedAt":"2026-10-05T21:21:08.839Z","media":{"type":"image","src":"/benchmarks/media/aafdc6200a414ae9.png","alt":"Ling 3.1 Flash: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-05T21:21:08.839Z","quality":{"score":37.9,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Central placement, separate airborne black paint and the elongated crimson coating earn substantial state and placement credit. Cyan contact remains unresolved. A few isolated sphere images provide limited optical evidence, while the defining repeated mirror structure is missing. Readable opaque metallic volume and restrained paint scale justify increased material credit. Core: 61.5/170; including the floor reward: 61.5/180.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":22,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":13.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":18,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0023316062176165792,"y":0.00621761658031088,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"inclusionai-ling-3-1-flash-svg-ground-truth-7c88674d15db6222","modelId":"inclusionai/ling-3.1-flash","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T17:19:36.688Z","media":{"type":"image","src":"/benchmarks/media/71245ee73be71e04.png","alt":"Ling 3.1 Flash: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T17:19:36.688Z","quality":{"score":30.6,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are partly correct."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"space-bunny-alpha-svg-ground-truth-7c88674d15db6222-from-20260929","modelId":"space-bunny-alpha","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T17:57:55.719Z","edition":{"id":"from-20260929","label":"29 Sep onward"},"media":{"type":"image","src":"/benchmarks/media/39f72a0b69066fd1.png","alt":"Space Bunny Alpha: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T17:57:55.719Z","quality":{"score":42.7,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-3-8-flash-svg-ground-truth-7c88674d15db6222","modelId":"google/gemini-3.8-flash","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:09.419Z","media":{"type":"image","src":"/benchmarks/media/aae27d86230a59d3.png","alt":"Gemini 3.8 Flash: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:09.419Z","quality":{"score":42.3,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are partly correct."},{"label":"Physical state","text":"The physical state is partly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"x-ai-grok-4-7-svg-ground-truth-7c88674d15db6222","modelId":"x-ai/grok-4.7","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:09.750Z","media":{"type":"image","src":"/benchmarks/media/7052cc71e7b684c5.png","alt":"Grok 4.7: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:09.750Z","quality":{"score":33.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"fledge-alpha-svg-ground-truth-7c88674d15db6222","modelId":"fledge-alpha","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:09.248Z","media":{"type":"image","src":"/benchmarks/media/7c6aab896b058721.png","alt":"Fledge Alpha: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:09.248Z","quality":{"score":36.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-svg-ground-truth-7c88674d15db6222","modelId":"deepseek/deepseek-v4.1-flash","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:08.751Z","media":{"type":"image","src":"/benchmarks/media/ac18776076188458.png","alt":"DeepSeek-V4.1-Flash: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:08.751Z","quality":{"score":39.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-opus-5-5-svg-ground-truth-7c88674d15db6222","modelId":"anthropic/claude-opus-5.5","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:08.408Z","media":{"type":"image","src":"/benchmarks/media/40fbe76e3262e69e.png","alt":"Claude Opus 5.5: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:08.408Z","quality":{"score":80.6,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is convincingly reconstructed."},{"label":"Lighting and surfaces","text":"Lighting and surface response are partly consistent."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are convincing and well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are mostly well controlled."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":-0.008841621501232716,"y":-0.0002787353996193628,"scale":0.9999999999999996},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-svg-ground-truth-7c88674d15db6222","modelId":"moonshotai/kimi-k3","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:10.108Z","media":{"type":"image","src":"/benchmarks/media/3899e2e3371b3b9c.png","alt":"Kimi K3: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:10.108Z","quality":{"score":40,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-1-3-svg-ground-truth-7c88674d15db6222","modelId":"meta/muse-spark-1.3","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:19:00.165Z","media":{"type":"image","src":"/benchmarks/media/d2e00fe0912e3024.png","alt":"Muse Spark 1.3: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:19:00.165Z","quality":{"score":28.6,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are partly correct."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"union-alpha-svg-ground-truth-7c88674d15db6222","modelId":"union-alpha","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:22:08.527Z","media":{"type":"image","src":"/benchmarks/media/1d44ce5401016e69.png","alt":"Union Alpha: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:22:08.527Z","quality":{"score":63.3,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-pro-0813-svg-ground-truth-7c88674d15db6222","modelId":"deepseek/deepseek-v4-pro-0813","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:08.906Z","media":{"type":"image","src":"/benchmarks/media/749e0e6fed7c74a2.png","alt":"DeepSeek-V4-Pro-0813: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:08.906Z","quality":{"score":39.6,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-luna-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-5.6-luna","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:23:13.060Z","media":{"type":"image","src":"/benchmarks/media/249455285d592478.png","alt":"GPT-5.6 Luna: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:23:13.060Z","quality":{"score":32.7,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are partly correct."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-sol-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-5.6-sol","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:25:08.869Z","media":{"type":"image","src":"/benchmarks/media/c2715a080833a3a7.png","alt":"GPT-5.6 Sol: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:25:08.869Z","quality":{"score":43.5,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-luna-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-6-luna","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:26:34.611Z","media":{"type":"image","src":"/benchmarks/media/97070ba02a603a7e.png","alt":"GPT-6 Luna: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:26:34.611Z","quality":{"score":30.3,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"nvidia-nemotron-3-ultra-550b-a55b-svg-ground-truth-7c88674d15db6222","modelId":"nvidia/nemotron-3-ultra-550b-a55b","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:31:04.133Z","media":{"type":"image","src":"/benchmarks/media/c7136f740d8ce893.png","alt":"Nemotron 3 Ultra: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:31:04.133Z","quality":{"score":20.5,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"owl-alpha-svg-ground-truth-7c88674d15db6222","modelId":"owl-alpha","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:32:42.903Z","media":{"type":"image","src":"/benchmarks/media/4d4c5fd329d38641.png","alt":"Owl Alpha: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:32:42.903Z","quality":{"score":12.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are poorly resolved."},{"label":"Colour and finish","text":"Colour and surface finish are poorly resolved."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-sol-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-6-sol","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:30:30.838Z","media":{"type":"image","src":"/benchmarks/media/17d46fd773486696.png","alt":"GPT-6 Sol: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:30:30.838Z","quality":{"score":50.7,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-fable-5-1-svg-ground-truth-7c88674d15db6222","modelId":"anthropic/claude-fable-5.1","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:08.214Z","media":{"type":"image","src":"/benchmarks/media/ee9d917d458961c2.png","alt":"Claude Fable 5.1: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:08.214Z","quality":{"score":64.1,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-6-terra-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-5.6-terra","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:25:19.399Z","media":{"type":"image","src":"/benchmarks/media/6a9a4cc19a8c35ce.png","alt":"GPT-5.6 Terra: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:25:19.399Z","quality":{"score":36.7,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-35b-a3b-svg-ground-truth-7c88674d15db6222","modelId":"qwen/qwen3.6-35b-a3b","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:39:04.729Z","media":{"type":"image","src":"/benchmarks/media/583caa6ded0b5b66.png","alt":"Qwen3.6-35B-A3B: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:39:04.729Z","quality":{"score":12.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-astra-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-6-astra","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:26:27.382Z","media":{"type":"image","src":"/benchmarks/media/740d6287d6078fc3.png","alt":"GPT-6 Astra: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:26:27.382Z","quality":{"score":69.2,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response are partly consistent."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are convincing and well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail show uneven control."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"mistralai-mistral-large-4-0-svg-ground-truth-7c88674d15db6222","modelId":"mistralai/mistral-large-4-0","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:32:04.160Z","media":{"type":"image","src":"/benchmarks/media/9999b4edce9881d6.png","alt":"Mistral Large 4: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:32:04.160Z","quality":{"score":38.1,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-max-0902-svg-ground-truth-7c88674d15db6222","modelId":"qwen/qwen3.8-max-0902","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:22:04.064Z","media":{"type":"image","src":"/benchmarks/media/fad983345b3ba2ee.png","alt":"Qwen3.8-Max-0902: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:22:04.064Z","quality":{"score":61.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-27b-svg-ground-truth-7c88674d15db6222","modelId":"qwen/qwen3.8-27B","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:40:31.571Z","media":{"type":"image","src":"/benchmarks/media/5e5876c29e9825cf.png","alt":"Qwen3.8-27B: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:40:31.571Z","quality":{"score":42.1,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-sonnet-5-5-svg-ground-truth-7c88674d15db6222","modelId":"anthropic/claude-sonnet-5.5","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:08.595Z","media":{"type":"image","src":"/benchmarks/media/92f33967645727dd.png","alt":"Claude Sonnet 5.5: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:08.595Z","quality":{"score":79.5,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is convincingly reconstructed."},{"label":"Lighting and surfaces","text":"Lighting and surface response are partly consistent."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail show uneven control."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-6-1-sol-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-6.1-sol","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:30:55.299Z","media":{"type":"image","src":"/benchmarks/media/1837668e6382fbd6.png","alt":"GPT-6.1 Sol: Ground Truth SVG benchmark","thumbnail":"/assets/homepage/thumbnails/bbeb4bf66d615544519f2d3f.png"},"firstGeneratedAt":"2026-10-06T18:30:55.299Z","quality":{"score":67.3,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are convincing and well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-svg-ground-truth-7c88674d15db6222","modelId":"minimax/minimax-m3","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:18:24.532Z","media":{"type":"image","src":"/benchmarks/media/7e5de8bac1c4cbb4.png","alt":"MiniMax M3: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:18:24.532Z","quality":{"score":28.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0.18502062122426036,"y":0.1896110592128085,"scale":0.7408182206817179},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"z-ai-glm-5-3-svg-ground-truth-7c88674d15db6222","modelId":"z-ai/glm-5.3","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:15:09.588Z","media":{"type":"image","src":"/benchmarks/media/2a7cd226ef0a8070.png","alt":"GLM-5.3: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:15:09.588Z","quality":{"score":49.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response are partly consistent."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail show uneven control."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"ox-alpha-svg-ground-truth-7c88674d15db6222","modelId":"ox-alpha","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:21:14.586Z","media":{"type":"image","src":"/benchmarks/media/a8f4fd40f8318a10.png","alt":"Ox Alpha: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:21:14.586Z","quality":{"score":63.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is convincingly reconstructed."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-flash-svg-ground-truth-7c88674d15db6222","modelId":"xiaomi/mimo-v2.6-flash","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:17:12.455Z","media":{"type":"image","src":"/benchmarks/media/de32a1168f23ded4.png","alt":"MiMo-V2.6-Flash: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:17:12.455Z","quality":{"score":50,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response are partly consistent."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"tencent-hy4-preview-svg-ground-truth-7c88674d15db6222","modelId":"tencent/hy4-preview","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T19:45:40.328Z","media":{"type":"image","src":"/benchmarks/media/37e9826ed925adbd.png","alt":"Hy4 preview: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T19:45:40.328Z","quality":{"score":41.3,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"mistralai-mistral-large-4-0-svg-pelican-f998b992640eff5e","modelId":"mistralai/mistral-large-4-0","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-10-06","generatedAt":"2026-10-06T20:09:38.503Z","media":{"type":"image","src":"/benchmarks/media/105b401c489939be.png","alt":"Mistral Large 4: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-06T20:09:38.503Z","quality":{"score":70.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A coherent, appealing pelican illustration with a continuous body silhouette, readable bicycle and grounded wheels. Simplified legs and ambiguous steering and pedal connections limit construction. Clean shapes and expressive accessories succeed, while shallow depth, modest lighting and sparse material differentiation leave substantial room for refinement.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"mistralai-mistral-large-4-0-svg-raccoon-7e17566f15394e20","modelId":"mistralai/mistral-large-4-0","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-06","generatedAt":"2026-10-06T20:09:38.520Z","media":{"type":"image","src":"/benchmarks/media/9d7d9e801311ab49.png","alt":"Mistral Large 4: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-06T20:09:38.520Z","quality":{"score":69.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The central interaction is clear, with recognizable animals, expressive faces and a coherent sunset composition. Clean contours and readable accessories give the cartoon polish. Simplified joints, uncertain steering attachment, shallow hull volume and limited directional lighting constrain its physical resolution.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"mistralai-mistral-large-4-0-svg-beach-48eb1c58439ea576","modelId":"mistralai/mistral-large-4-0","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-10-06","generatedAt":"2026-10-06T20:09:38.535Z","media":{"type":"image","src":"/benchmarks/media/59d9c817f79fc85a.png","alt":"Mistral Large 4: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-06T20:09:38.535Z","quality":{"score":65.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A clean, appealing beach illustration with clear subject coverage, a cohesive sunset palette and expressive animal details. Bodies and accessories resolve basic construction coherently, although joints remain schematic. Layered shapes establish modest depth; generic oval shadows and flat surfaces limit physical richness. Repeated forms and sparse articulation leave room for more accomplished drawing.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"mistralai-mistral-large-4-0-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"mistralai/mistral-large-4-0","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-06","generatedAt":"2026-10-06T20:09:38.550Z","media":{"type":"image","src":"/benchmarks/media/65b1cc17f081c9b9.png","alt":"Mistral Large 4: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-06T20:09:38.550Z","quality":{"score":28.8,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The central sphere, airborne black and deformed crimson support substantial state and placement credit. Recent cyan contact and crimson transport remain uncertain. Sparse wall patches and glossy gradients provide little evidence of the required extensive mirror system, while the missing centerline lamp and bulky crimson remain visible weaknesses.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":17,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":11,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":12,"max":120},{"id":"materialsScale","label":"Materials and scale","score":5.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.004961411245865491,"y":0.009095920617420067,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"mistralai-mistral-large-4-0-svg-mona-lisa-63163c6b35a3998f","modelId":"mistralai/mistral-large-4-0","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-10-06","generatedAt":"2026-10-06T21:03:01.871Z","media":{"type":"image","src":"/benchmarks/media/5b0d3751561378b5.png","alt":"Mistral Large 4: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-06T21:03:01.871Z","quality":{"score":41.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-07","rationale":"A recognizable, cleanly rendered interpretation with a restrained palette and some facial modeling. Its enlarged frontal head, simplified expression, compressed low hands and schematic landscape substantially limit likeness and construction. The smooth rendering does not reproduce the painting's nuanced anatomy, clothing or atmospheric depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.055065368765061164,"y":0.2163442635236202,"scale":1.7949909856398993},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"mistralai-mistral-large-4-0-svg-selfie-6ef92765195fa84b","modelId":"mistralai/mistral-large-4-0","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A heavy ginger cat stays planted on a terrace as wind carries papers around it. The scene tests weight, expression and the contrast between stable and moving forms.","promptHash":"6ef92765195fa84b","date":"2026-10-06","generatedAt":"2026-10-06T22:52:57.688Z","media":{"type":"image","src":"/benchmarks/media/a586afb025909f47.png","alt":"Mistral Large 4: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-06T22:52:57.688Z","quality":{"score":72.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"An appealing, coherent illustration whose enormous rounded cat anchors the windblown composition. Warm shading and floating sketches communicate the brief well. Simplified paws, faint ear construction, inconsistent wind cues and limited material texture keep the execution below exceptional.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A heavy ginger cat stays planted on a terrace as wind carries papers around it. The scene tests weight, expression and the contrast between stable and moving forms.","promptSummary":"A heavy ginger cat stays planted on a terrace as wind carries papers around it. The scene tests weight, expression and the contrast between stable and moving forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-8-flash-svg-ground-truth-7c88674d15db6222","modelId":"qwen/qwen3.8-flash","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T18:21:58.781Z","media":{"type":"image","src":"/benchmarks/media/d9a1b1766aed4eef.png","alt":"Qwen3.8-Flash: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T18:21:58.781Z","quality":{"score":37,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-5-9b-svg-ground-truth-7c88674d15db6222","modelId":"qwen/qwen3.5-9b","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T23:23:59.907Z","media":{"type":"image","src":"/benchmarks/media/a17c4d43912d973e.png","alt":"Qwen3.5-9B: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T23:23:59.907Z","quality":{"score":12,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"openai-gpt-5-5-svg-ground-truth-7c88674d15db6222","modelId":"openai/gpt-5.5","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T23:28:25.210Z","media":{"type":"image","src":"/benchmarks/media/015188b43bcb4310.png","alt":"GPT-5.5: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T23:28:25.210Z","quality":{"score":40.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-07","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"aleph-alpha-kolibri-1-svg-ground-truth-7c88674d15db6222","modelId":"Aleph-Alpha/Kolibri-1","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-06","generatedAt":"2026-10-06T23:37:07.900Z","media":{"type":"image","src":"/benchmarks/media/46715071bc1dd559.png","alt":"Kolibri-1: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-06T23:37:07.900Z","quality":{"score":21.2,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-haiku-5-5-svg-mona-lisa-e7b85d8cd393c26f","modelId":"anthropic/claude-haiku-5.5","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-08","generatedAt":"2026-10-08T03:55:27.984Z","media":{"type":"image","src":"/benchmarks/media/7b053a38aae50d5a.png","alt":"Claude Haiku 5.5: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-08T03:42:17.832Z","quality":{"score":38.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A recognizable but basic Mona Lisa reinterpretation. Clean contours and smooth facial lighting provide coherence and volume. Replication remains limited by the frontal pose, elongated neck, schematic expression, unresolved crossed hands and sparse landscape. Lighting is more successful than anatomical construction and purposeful detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05248193601358682,"y":0.1778590962186721,"scale":1.4333294145603392},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-haiku-5-5-svg-raccoon-b4cedf42a4a26a4b","modelId":"anthropic/claude-haiku-5.5","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-08","generatedAt":"2026-10-08T03:55:27.984Z","media":{"type":"image","src":"/benchmarks/media/d46b766c75613b10.png","alt":"Claude Haiku 5.5: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-08T03:42:17.832Z","quality":{"score":47.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A tidy, appealing cartoon with recognizable animals and a pleasing sunset palette. Steering lacks visible controls, while schematic limbs and uncertain raccoon support weaken the requested interaction. Shallow vehicle construction, flat character lighting and modest displacement limit physical coherence. Clean shapes and readable identifying details remain strengths.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-haiku-5-5-svg-selfie-64223bae34ab729d","modelId":"anthropic/claude-haiku-5.5","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptHash":"64223bae34ab729d","date":"2026-10-08","generatedAt":"2026-10-08T03:55:27.983Z","media":{"type":"image","src":"/benchmarks/media/ad709960fd39387d.png","alt":"Claude Haiku 5.5: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-08T03:42:17.831Z","quality":{"score":46.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A coherent coral squid emblem with a compact mantle, restrained face and clean centered presentation. Its radial silhouette and floating placement work, but sharp arm tips, crowded roots and sparse surface modeling limit softness, construction and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":3.5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummary":"A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-haiku-5-5-svg-pelican-b618cda0f539e818","modelId":"anthropic/claude-haiku-5.5","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-08","generatedAt":"2026-10-08T03:55:27.984Z","media":{"type":"image","src":"/benchmarks/media/066a8724b76e085c.png","alt":"Claude Haiku 5.5: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-08T03:55:27.984Z","quality":{"score":61.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A cheerful, recognizable illustration with clean contours and a balanced warm palette. Aligned wheels and readable frame geometry establish the bicycle, but shoulder construction, feet and steering contact remain imprecise. Sparse drivetrain detail, flat materials and absent grounding shadows limit anatomical and spatial conviction.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-haiku-5-5-svg-beach-d030aa6f45fe19e9","modelId":"anthropic/claude-haiku-5.5","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-08","generatedAt":"2026-10-08T03:55:27.984Z","media":{"type":"image","src":"/benchmarks/media/915487381bc252d4.png","alt":"Claude Haiku 5.5: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-08T03:55:27.984Z","quality":{"score":70.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-20","rationale":"A cheerful, coherent beach illustration with attractive sunset colors, readable accessories and clean silhouettes. Its sun and sailboat reflections establish useful physical relationships. Simple animal construction, flattened foreground placement, sparse directional shadows and minimally differentiated surfaces limit its drawing refinement.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-haiku-5-5-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"anthropic/claude-haiku-5.5","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-08","generatedAt":"2026-10-08T03:55:27.984Z","media":{"type":"image","src":"/benchmarks/media/71a4ef75df50eb5f.png","alt":"Claude Haiku 5.5: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-08T03:55:27.984Z","quality":{"score":40.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The centered sphere, compact airborne black blob and differentiated cyan/crimson interaction support substantial state and placement credit. The few recognizable sphere and panel images establish limited optical relationships, but most of the required mirror system is missing. Rounded volume and localized paint warrant substantial material credit despite the soft, illustrative finish.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":25,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":8,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":24.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"anthropic-claude-haiku-5-5-svg-ground-truth-7c88674d15db6222","modelId":"anthropic/claude-haiku-5.5","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-08","generatedAt":"2026-10-08T04:01:49.536Z","media":{"type":"image","src":"/benchmarks/media/5a291cb08e1f55c8.png","alt":"Claude Haiku 5.5: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-08T04:01:49.536Z","quality":{"score":49.2,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"glyph-cluster-svg-selfie-c23df51df11ed0b1","modelId":"glyph-cluster","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"Punctuation-shaped companions gather inside a curled paper home. The scene tests expressive silhouettes, close interaction and the coherence of a detailed miniature environment.","promptHash":"c23df51df11ed0b1","date":"2026-10-08","generatedAt":"2026-10-08T16:31:10.149Z","media":{"type":"image","src":"/benchmarks/media/7880a7fdac7bbe68.png","alt":"Glyph Cluster: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-08T16:31:10.149Z","quality":{"score":32.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A charming palette and three expressive companions make the concept recognizable. The dominant paper enclosure overwhelms the small group, the plum character lacks a clear brace silhouette, and the cuddle remains difficult to trace. Sparse furnishings, ambiguous folds and generalized shading limit the promised lived-in richness and tactile depth.","categories":[{"id":"fidelity","label":"Task & likeness","score":3.5},{"id":"structure","label":"Anatomy & construction","score":2.5},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":3.5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"Punctuation-shaped companions gather inside a curled paper home. The scene tests expressive silhouettes, close interaction and the coherence of a detailed miniature environment.","promptSummary":"Punctuation-shaped companions gather inside a curled paper home. The scene tests expressive silhouettes, close interaction and the coherence of a detailed miniature environment.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"glyph-cluster-svg-mona-lisa-63163c6b35a3998f","modelId":"glyph-cluster","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-10-08","generatedAt":"2026-10-08T17:25:04.432Z","media":{"type":"image","src":"/benchmarks/media/53bad8841d81928d.png","alt":"Glyph Cluster: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-08T17:25:04.432Z","quality":{"score":11.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The portrait has a continuous body silhouette, attached sleeves and hands, and simple landscape depth. Its oversized pale head enclosure nearly erases the expression, while separated fists, ochre clothing and frontal construction weaken Mona Lisa likeness. Sparse detail, flat materials and ambiguous facial layers leave the drawing seriously unresolved.","categories":[{"id":"fidelity","label":"Task & likeness","score":2},{"id":"structure","label":"Anatomy & construction","score":0.5},{"id":"space","label":"Space & placement","score":2},{"id":"light","label":"Light & reflections","score":1},{"id":"detail","label":"Purposeful detail","score":1},{"id":"artistry","label":"Artistry & character","score":0.5},{"id":"effects","label":"Materials & effects","score":1},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"glyph-cluster-svg-beach-48eb1c58439ea576","modelId":"glyph-cluster","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-10-08","generatedAt":"2026-10-08T17:25:04.433Z","media":{"type":"image","src":"/benchmarks/media/f72d7dfcd65a0932.png","alt":"Glyph Cluster: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-08T17:25:04.433Z","quality":{"score":46.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A cheerful, recognizable beach illustration with playful accessories and mostly clean silhouettes. Sunset colors and soft glow supply atmosphere, but abbreviated anatomy, unexplained airborne animals, weak ground contact and absent water reflections limit physical coherence. Small details add interest without consistently resolving construction or materials.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"glyph-cluster-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"glyph-cluster","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-08","generatedAt":"2026-10-08T17:25:04.432Z","media":{"type":"image","src":"/benchmarks/media/da50f619d3e19694.png","alt":"Glyph Cluster: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-08T17:25:04.432Z","quality":{"score":12.8,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Rounded sphere volume, compact colored forms and crimson-below-cyan ordering provide limited agreement. The visible paint does not establish the required interaction history, the sphere and black blob are displaced from the prescribed projection, and most of the defining mirror-room structure is missing or unresolved. The faint reflection echoes earn limited optical credit.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":3,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":8,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0037432961799217243,"y":0.05381505281183083,"scale":1.161834242728283},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"glyph-cluster-svg-ground-truth-7c88674d15db6222","modelId":"glyph-cluster","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-08","generatedAt":"2026-10-08T16:31:10.149Z","media":{"type":"image","src":"/benchmarks/media/e451c210da58acab.png","alt":"Glyph Cluster: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-08T16:31:10.149Z","quality":{"score":19.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are partly correct."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"glyph-cluster-svg-raccoon-7e17566f15394e20","modelId":"glyph-cluster","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-08","generatedAt":"2026-10-08T16:31:10.149Z","media":{"type":"image","src":"/benchmarks/media/2a5c4b817923f9b9.png","alt":"Glyph Cluster: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-08T16:31:10.149Z","quality":{"score":42.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"Recognizable animals and clean shapes give the scene charm, but the requested steering action is visibly incomplete. Simplified joints, weak vehicle construction and uncertain water contact limit physical coherence. Flat lighting, sparse functional detail and a small focal group keep the illustration basic despite its tidy edges.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":2.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":3},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"glyph-cluster-svg-pelican-f998b992640eff5e","modelId":"glyph-cluster","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-10-08","generatedAt":"2026-10-08T16:31:10.149Z","media":{"type":"image","src":"/benchmarks/media/847f257ff57455ea.png","alt":"Glyph Cluster: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-08T16:31:10.149Z","quality":{"score":44.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A readable cartoon with open framing, aligned bicycle wheels and modest atmospheric effects. Pelican likeness is approximate, while the narrow neck, crowded feet, absent steering contact and unclear drivetrain weaken the riding action. Clean shapes help presentation, but shallow space and inconsistent shading limit the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":3.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-27b-svg-ground-truth-7c88674d15db6222","modelId":"qwen/qwen3.6-27b","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-08","generatedAt":"2026-10-08T18:17:47.202Z","media":{"type":"image","src":"/benchmarks/media/3d08813815cbff51.png","alt":"Qwen3.6-27B: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-08T18:17:47.202Z","quality":{"score":26.7,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-08","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are partly correct."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-27b-svg-selfie-aea705fcb6511459","modelId":"qwen/qwen3.6-27b","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptHash":"aea705fcb6511459","date":"2026-10-08","generatedAt":"2026-10-08T18:17:47.202Z","media":{"type":"image","src":"/benchmarks/media/a6596f4f29c6623a.png","alt":"Qwen3.6-27B: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-08T18:17:47.202Z","quality":{"score":56.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"An appealing decorative illustration with coordinated ornament, flowing panels and three glowing orbital lights. The arm arrangement reverses the requested sides. Uncertain neck continuity, concealed leg articulation, simplified paws, crowded chest jewelry and shallow surface modeling limit pose clarity and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummary":"A floating capybara in ornate robes tests character anatomy, layered fabric and the placement of surrounding luminous forms.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-27b-svg-mona-lisa-e7b85d8cd393c26f","modelId":"qwen/qwen3.6-27b","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-08","generatedAt":"2026-10-08T18:17:47.202Z","media":{"type":"image","src":"/benchmarks/media/edaf8e3dbb5cf347.png","alt":"Qwen3.6-27B: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-08T18:17:47.202Z","quality":{"score":30.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"The portrait loosely evokes Mona Lisa through its palette and silhouette. Readable pupils, brows and lips provide limited detail, and the torso remains continuous. The frontal elongated face, unresolved chest-level limb forms, shallow landscape and pervasive blur substantially weaken likeness, anatomy and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":2.5},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":2.5},{"id":"effects","label":"Materials & effects","score":3},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05895868864326423,"y":0.3365506339073581,"scale":2.117000016612674},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-27b-svg-raccoon-b4cedf42a4a26a4b","modelId":"qwen/qwen3.6-27b","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-08","generatedAt":"2026-10-08T18:17:47.202Z","media":{"type":"image","src":"/benchmarks/media/59c5afac3e074ccc.png","alt":"Qwen3.6-27B: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-08T18:17:47.202Z","quality":{"score":45.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"A readable cartoon with balanced colors, clean shapes and a clearly supported raccoon. Continuous body construction and modest shading deserve credit. Rudimentary limbs, unresolved steering connections, stiff gestures and schematic water effects limit physical credibility and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":2.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-27b-svg-beach-d030aa6f45fe19e9","modelId":"qwen/qwen3.6-27b","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-08","generatedAt":"2026-10-08T18:17:47.202Z","media":{"type":"image","src":"/benchmarks/media/7549f8636357c41f.png","alt":"Qwen3.6-27B: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-08T18:17:47.202Z","quality":{"score":56.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A recognizable sunset beach with balanced palms, clean shapes and effective warm–cool contrast. Small animals and play equipment address the brief, but rudimentary anatomy, uncertain water contacts, sparse reflections and flat materials keep the drawing basic or uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-27b-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"qwen/qwen3.6-27b","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-08","generatedAt":"2026-10-08T18:17:47.202Z","media":{"type":"image","src":"/benchmarks/media/bd9dd016ef549ec3.png","alt":"Qwen3.6-27B: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-08T18:17:47.202Z","quality":{"score":15.5,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The artwork depicts a shaded sphere, distinct paint colors and a separate airborne black blob. Crimson's apex overlap does not establish its earlier interaction and transport, while cyan's recent contact remains unsupported. Horizontal centering supplies limited camera agreement, but the sphere, black blob and prominent panel have substantial placement or projection departures. A few planar echoes provide reflection evidence; isolated patches and gradients replace most of the required mirror structure.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":8,"max":120},{"id":"materialsScale","label":"Materials and scale","score":4,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.00024751824801330907,"y":0.08833887382986909,"scale":0.7408182206817179},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"qwen-qwen3-6-27b-svg-pelican-b618cda0f539e818","modelId":"qwen/qwen3.6-27b","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-08","generatedAt":"2026-10-08T18:17:47.202Z","media":{"type":"image","src":"/benchmarks/media/92854c7671489a77.png","alt":"Qwen3.6-27B: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-08T18:17:47.202Z","quality":{"score":58.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-08","rationale":"A lively, coherent illustration with recognizable pelican features, purposeful accessories and useful surface variation. Ambiguous neck layering, uncertain pedal engagement and an unresolved drivetrain limit construction. Aligned wheels and layered scenery establish placement, while inconsistent shadow direction and crowded overlaps constrain finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"microsoft-mai-thinking-1-svg-ground-truth-7c88674d15db6222","modelId":"microsoft/mai-thinking-1","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-09","generatedAt":"2026-10-09T01:56:20.220Z","media":{"type":"image","src":"/benchmarks/media/ea3bd6787e36fe61.png","alt":"MAI-Thinking-1: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-09T01:56:20.220Z","quality":{"score":1.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-09","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are poorly resolved."},{"label":"Colour and finish","text":"Colour and surface finish are poorly resolved."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"microsoft-mai-thinking-1-svg-mona-lisa-63163c6b35a3998f","modelId":"microsoft/mai-thinking-1","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-10-09","generatedAt":"2026-10-09T01:56:20.220Z","media":{"type":"image","src":"/benchmarks/media/62158afe867066d6.png","alt":"MAI-Thinking-1: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-09T01:56:20.220Z","quality":{"score":11.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A recognizable Mona Lisa motif emerges from dark hair, a pale face and a landscape band. Replication remains weak: the oversized frontal head, broad smile, unresolved torso and missing crossed-hand anatomy diverge substantially from the reference. Simple fills and sparse strokes provide limited volume, material definition or painterly complexity.","categories":[{"id":"fidelity","label":"Task & likeness","score":1},{"id":"structure","label":"Anatomy & construction","score":0.5},{"id":"space","label":"Space & placement","score":1},{"id":"light","label":"Light & reflections","score":1.5},{"id":"detail","label":"Purposeful detail","score":1},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":0.5},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.058125998811908386,"y":0.28852661657037015,"scale":1.5683121854901685},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"microsoft-mai-thinking-1-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"microsoft/mai-thinking-1","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-09","generatedAt":"2026-10-09T01:56:20.220Z","media":{"type":"image","src":"/benchmarks/media/88b2bb766d8c60a5.png","alt":"MAI-Thinking-1: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-09T01:56:20.220Z","quality":{"score":8,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The small paint features and convex shading deserve moderate material credit. The sphere remains undersized and low, with black unusually close to its apex. Crimson and cyan do not establish the required impact histories. Faint reflected echoes supply limited optical evidence, while the defining recursive room, nested sphere and emitter structures remain unestablished.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":3,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":2.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":4,"max":120},{"id":"materialsScale","label":"Materials and scale","score":2,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0013976311412203539,"y":0.04421502248986251,"scale":0.6376281516217733},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"microsoft-mai-thinking-1-svg-beach-48eb1c58439ea576","modelId":"microsoft/mai-thinking-1","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-10-09","generatedAt":"2026-10-09T01:56:20.220Z","media":{"type":"image","src":"/benchmarks/media/30b927f59add9e7d.png","alt":"MAI-Thinking-1: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-09T01:56:20.220Z","quality":{"score":44.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A coherent, cheerful beach scene with balanced placement, clean shapes and controlled sunset colours. Animals, birds and playful accessories are recognizable, but basic dolphin construction, weak grounding, schematic reflections and sparse material detail limit the drawing's accomplishment.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":3.5},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"microsoft-mai-thinking-1-svg-raccoon-7e17566f15394e20","modelId":"microsoft/mai-thinking-1","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-09","generatedAt":"2026-10-09T01:56:20.220Z","media":{"type":"image","src":"/benchmarks/media/32b40e0512468675.png","alt":"MAI-Thinking-1: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-09T01:56:20.220Z","quality":{"score":37.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The recognizable raccoon and clean, shaded forms make a readable cartoon. The dog's steering remains undemonstrated, while ambiguous seating contacts and a minimally constructed craft weaken the central interaction. Controlled gradients and edges deserve credit, but anatomy, purposeful detail and water displacement remain basic.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":2.5},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":3.5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"microsoft-mai-thinking-1-svg-selfie-790b8460d788f948","modelId":"microsoft/mai-thinking-1","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A fluffy moth tends a lantern greenhouse amid a gardening mishap. The scene tests insect anatomy, expressive action, translucent wings and the interaction of light with small objects.","promptHash":"790b8460d788f948","date":"2026-10-09","generatedAt":"2026-10-09T01:56:20.219Z","media":{"type":"image","src":"/benchmarks/media/80cb224cd0142e91.png","alt":"MAI-Thinking-1: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-09T01:56:20.219Z","quality":{"score":46.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A recognizable lantern-and-moth illustration with appealing coloured wings and tidy outer contours. Simple construction, a small focal subject and crowded overlaps limit character and depth. The face, six-leg anatomy, seedling and antenna-driven watering joke remain insufficiently readable.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A fluffy moth tends a lantern greenhouse amid a gardening mishap. The scene tests insect anatomy, expressive action, translucent wings and the interaction of light with small objects.","promptSummary":"A fluffy moth tends a lantern greenhouse amid a gardening mishap. The scene tests insect anatomy, expressive action, translucent wings and the interaction of light with small objects.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"microsoft-mai-thinking-1-svg-pelican-f998b992640eff5e","modelId":"microsoft/mai-thinking-1","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-10-09","generatedAt":"2026-10-09T01:56:20.220Z","media":{"type":"image","src":"/benchmarks/media/643a89b90e61e83d.png","alt":"MAI-Thinking-1: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-09T01:56:20.220Z","quality":{"score":56.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A recognizable pelican and bicycle combine clean rounded forms, expressive facial details and effective surface shading. Aligned wheels and soft shadows establish placement, but unresolved pedal engagement, absent steering contact and sparse drivetrain construction weaken the riding action. Limited feather complexity and shallow space constrain an otherwise appealing illustration.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-31b-it-svg-pelican-b618cda0f539e818","modelId":"google/gemma-4-31b-it","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-09","generatedAt":"2026-10-09T14:37:49.765Z","media":{"type":"image","src":"/benchmarks/media/a78655f12aec6d0d.png","alt":"Gemma 4 31B: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-09T14:37:49.765Z","quality":{"score":37.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"A clean, readable illustration with balanced placement and coherent basic geometry. The sky gradient and ground shadow establish diffuse illumination. Primitive shapes and sparse detail limit artistry, while unclear leg attachment, missing pedal contact and unresolved steering weaken the riding action.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-31b-it-svg-raccoon-b4cedf42a4a26a4b","modelId":"google/gemma-4-31b-it","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-09","generatedAt":"2026-10-09T14:37:49.765Z","media":{"type":"image","src":"/benchmarks/media/05ffd17dd2b475b6.png","alt":"Gemma 4 31B: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-09T14:37:49.765Z","quality":{"score":38.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"Clean, simple silhouettes communicate a raccoon riding on a dog. Limited animal specificity, absent steering controls and a flat platform weaken the requested scene. Small facial marks provide purposeful detail, while unresolved support, flat lighting and schematic water restrict physical coherence. Primitive but coherent construction supports basic drawing quality.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":1},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-26b-a4b-it-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"google/gemma-4-26b-a4b-it","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-09","generatedAt":"2026-10-09T15:02:58.763Z","media":{"type":"image","src":"/benchmarks/media/f475aebb180f92ec.png","alt":"Gemma 4 26B A4B: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:02:58.763Z","quality":{"score":10,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Separate airborne black and contact-compatible paint flattening earn partial state credit. Settlement, adhesion and differentiated paint transport remain unresolved. The enlarged, low sphere and thin lamp strip substantially depart from the required camera, while the defining mirror structure is missing. Weak material appearance and oversized black paint limit material credit. The retained base total is 18/170; the floor-damage reward is 0, giving 18/180.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":10,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":2,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":0,"max":120},{"id":"materialsScale","label":"Materials and scale","score":2.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0012071262719021622,"y":0.21275313939920185,"scale":1.6404982390570433},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-26b-a4b-it-svg-ground-truth-7c88674d15db6222","modelId":"google/gemma-4-26b-a4b-it","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-09","generatedAt":"2026-10-09T15:02:58.788Z","media":{"type":"image","src":"/benchmarks/media/8e8c89d54e3d87a6.png","alt":"Gemma 4 26B A4B: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:02:58.788Z","quality":{"score":14.2,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-09","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are poorly resolved."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-26b-a4b-it-svg-selfie-8870f065df3cd897","modelId":"google/gemma-4-26b-a4b-it","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small luminous creature proudly holding an empty glass jar tests expression, transparent materials and the balance of character and prop.","promptHash":"8870f065df3cd897","date":"2026-10-09","generatedAt":"2026-10-09T15:02:58.441Z","media":{"type":"image","src":"/benchmarks/media/342e8c1f698eafe4.png","alt":"Gemma 4 26B A4B: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:02:58.441Z","quality":{"score":52.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"An appealing, readable sparkle-creature with luminous colour and expressive eyes. Simple geometric construction limits sophistication, while sharp body points, ambiguous hand contact and a minimally resolved jar weaken the requested soft, plump character and treasure-presenting pose.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small luminous creature proudly holding an empty glass jar tests expression, transparent materials and the balance of character and prop.","promptSummary":"A small luminous creature proudly holding an empty glass jar tests expression, transparent materials and the balance of character and prop.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-26b-a4b-it-svg-mona-lisa-58920dfd672bf618","modelId":"google/gemma-4-26b-a4b-it","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-10-09","generatedAt":"2026-10-09T15:02:58.441Z","media":{"type":"image","src":"/benchmarks/media/ecde2e9f2e7d1095.png","alt":"Gemma 4 26B A4B: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:02:58.441Z","quality":{"score":17.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The centered dark-haired portrait provides a loose reference cue, but its elongated, nearly featureless face and undifferentiated torso fall far short of replication. Crossed hands, the subtle smile, clothing structure and landscape are unresolved. Clean silhouettes and soft gradients provide limited coherence.","categories":[{"id":"fidelity","label":"Task & likeness","score":0.5},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":1},{"id":"light","label":"Light & reflections","score":2},{"id":"detail","label":"Purposeful detail","score":1},{"id":"artistry","label":"Artistry & character","score":2},{"id":"effects","label":"Materials & effects","score":1},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.0028143507391553,"y":0.14206048168547697,"scale":1.1972173631218093},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-26b-a4b-it-svg-pelican-b618cda0f539e818","modelId":"google/gemma-4-26b-a4b-it","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-09","generatedAt":"2026-10-09T15:02:58.441Z","media":{"type":"image","src":"/benchmarks/media/44fe2fa71160cf5f.png","alt":"Gemma 4 26B A4B: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:02:58.441Z","quality":{"score":34,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"A clean, balanced icon suggests a pelican riding a bicycle. Continuous contours and restrained colors support its finish, but the abbreviated bill, undifferentiated body, unresolved riding contacts and incomplete bicycle mechanism limit likeness and construction. Sparse detail and flat lighting keep the drawing basic despite its clarity.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":3},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":2.5},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-26b-a4b-it-svg-raccoon-b4cedf42a4a26a4b","modelId":"google/gemma-4-26b-a4b-it","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-09","generatedAt":"2026-10-09T15:02:58.441Z","media":{"type":"image","src":"/benchmarks/media/15f40588fcec2555.png","alt":"Gemma 4 26B A4B: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:02:58.441Z","quality":{"score":40.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The animal stack is recognizable, and clean silhouettes support a basic illustration. Steering is not depicted, while the craft's construction and water contact remain unresolved. Sparse anatomical and material detail limit the finish. The sky gradient and glow provide atmospheric lighting, but the subjects lack corresponding modeling.","categories":[{"id":"fidelity","label":"Task & likeness","score":4.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":2.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-26b-a4b-it-svg-beach-d030aa6f45fe19e9","modelId":"google/gemma-4-26b-a4b-it","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-09","generatedAt":"2026-10-09T15:02:58.441Z","media":{"type":"image","src":"/benchmarks/media/9d49aef6c06881ce.png","alt":"Gemma 4 26B A4B: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:02:58.441Z","quality":{"score":28.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A clean, balanced sunset scene demonstrates the beach, birds and a centered reflection. Animals and play receive only minimal treatment. Coherent layers and controlled gradients support readability, but unresolved aquatic anatomy, sparse details and weak surface modeling keep the drawing basic.","categories":[{"id":"fidelity","label":"Task & likeness","score":3.5},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":3.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":1},{"id":"artistry","label":"Artistry & character","score":1.5},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-31b-it-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"google/gemma-4-31b-it","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-09","generatedAt":"2026-10-09T15:04:56.814Z","media":{"type":"image","src":"/benchmarks/media/e2f6048e0d534799.png","alt":"Gemma 4 31B: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:04:56.814Z","quality":{"score":16,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The sphere's volume, separate airborne black blob and vertical color ordering are readable. The large, low sphere, plain surrounding planes and absent nested optical imagery leave the central geometric and reflection requirements unsupported. Paint contact and transport remain unresolved. Material credit recognizes the visible convex shading, highlights and distinct surface colors while retaining deductions for the schematic finish and oversized blobs.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":11,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":3,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":0,"max":120},{"id":"materialsScale","label":"Materials and scale","score":5.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.004334326593795162,"y":0.13199920254391587,"scale":1.8221188003905082},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-31b-it-svg-mona-lisa-58920dfd672bf618","modelId":"google/gemma-4-31b-it","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-10-09","generatedAt":"2026-10-09T15:04:56.814Z","media":{"type":"image","src":"/benchmarks/media/756a08f675381335.png","alt":"Gemma 4 31B: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:04:56.814Z","quality":{"score":35.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"The portrait is recognizable through its silhouette, palette and smile, with clean execution of simple shapes. Replication remains weak: the face and pose differ markedly, hands become fingerless ovals, clothing lacks construction, and the landscape loses its atmospheric detail. Basic gradients provide limited volume.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":3.5},{"id":"detail","label":"Purposeful detail","score":2},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05755481787273816,"y":0.35582254007275044,"scale":2.0544332106438867},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-31b-it-svg-beach-d030aa6f45fe19e9","modelId":"google/gemma-4-31b-it","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-09","generatedAt":"2026-10-09T15:04:56.814Z","media":{"type":"image","src":"/benchmarks/media/a107033319e5be6e.png","alt":"Gemma 4 31B: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:04:56.814Z","quality":{"score":29.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A coherent beach illustration with clean silhouettes, a sweeping shoreline and an effective sunset palette. Requested subjects are recognizable, but primitive animal construction, sparse detail, unsupported palm placement and schematic reflections limit drawing sophistication and physical credibility.","categories":[{"id":"fidelity","label":"Task & likeness","score":5},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":2.5},{"id":"detail","label":"Purposeful detail","score":2},{"id":"artistry","label":"Artistry & character","score":2},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-31b-it-svg-ground-truth-7c88674d15db6222","modelId":"google/gemma-4-31b-it","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-09","generatedAt":"2026-10-09T15:04:56.814Z","media":{"type":"image","src":"/benchmarks/media/39fde9fd74b5006c.png","alt":"Gemma 4 31B: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:04:56.814Z","quality":{"score":14.1,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-09","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemma-4-31b-it-svg-selfie-8870f065df3cd897","modelId":"google/gemma-4-31b-it","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small luminous creature proudly holding an empty glass jar tests expression, transparent materials and the balance of character and prop.","promptHash":"8870f065df3cd897","date":"2026-10-09","generatedAt":"2026-10-09T15:04:56.814Z","media":{"type":"image","src":"/benchmarks/media/29dc149cc48fa784.png","alt":"Gemma 4 31B: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-09T15:04:56.814Z","quality":{"score":60.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-09","rationale":"A clean, appealing luminous mascot with coherent body continuity and a readable jar-hugging pose. Rounded highlights, smooth gradients and balanced framing work well. The teardrop silhouette, modest eyes and nearly even grin only partly satisfy the character brief; sparse jar construction and simple geometry limit sophistication.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small luminous creature proudly holding an empty glass jar tests expression, transparent materials and the balance of character and prop.","promptSummary":"A small luminous creature proudly holding an empty glass jar tests expression, transparent materials and the balance of character and prop.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"pixel-canary-svg-ground-truth-7c88674d15db6222","modelId":"pixel-canary","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-10","generatedAt":"2026-10-10T03:04:38.877Z","media":{"type":"image","src":"/benchmarks/media/43f54d542a5adc2f.png","alt":"Pixel Canary: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-10T03:04:38.877Z","quality":{"score":44.1,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-10","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is partly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"missingno-svg-mona-lisa-63163c6b35a3998f","modelId":"missingno","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-10-11","generatedAt":"2026-10-11T20:25:38.490Z","media":{"type":"image","src":"/benchmarks/media/f17b4507ca3ace44.png","alt":"MISSINGNO.: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-11T20:25:38.490Z","quality":{"score":28.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"Clean contours, connected body forms and gentle gradients support a basic, coherent portrait. Likeness remains weak: the face dominates the body, the seated angle is lost, and crossed hands become dark lobes without traceable wrists or fingers. Readable facial details add some drawing complexity, but clothing, landscape and material treatment remain sparse.","categories":[{"id":"fidelity","label":"Task & likeness","score":1.5},{"id":"structure","label":"Anatomy & construction","score":2.5},{"id":"space","label":"Space & placement","score":2.5},{"id":"light","label":"Light & reflections","score":3},{"id":"detail","label":"Purposeful detail","score":2},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.032998872842683595,"y":0.2303751072525351,"scale":1.8496565995583263},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"missingno-svg-beach-48eb1c58439ea576","modelId":"missingno","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-10-11","generatedAt":"2026-10-11T20:26:03.413Z","media":{"type":"image","src":"/benchmarks/media/a73763918907643a.png","alt":"MISSINGNO.: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-11T20:26:03.413Z","quality":{"score":43.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"A coherent sunset beach with clear requested subjects, balanced framing and clean shapes. The aligned solar reflection and warm gradients establish atmosphere. Primitive animal construction, ambiguous furniture, sparse depth cues and minimally developed surfaces limit its drawing complexity and physical richness.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":3.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"missingno-svg-pelican-f998b992640eff5e","modelId":"missingno","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-10-11","generatedAt":"2026-10-11T20:26:10.289Z","media":{"type":"image","src":"/benchmarks/media/22a90ed6b2594ec3.png","alt":"MISSINGNO.: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-11T20:26:10.289Z","quality":{"score":58.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"A coherent, cheerful illustration with recognizable pelican features, balanced composition, clean contours and useful small details. Its controlled curves and layered wheels demonstrate competent drawing skill. Ambiguous pedal and steering contacts, sparse drivetrain construction and weak directional lighting limit the riding mechanics and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"missingno-svg-ground-truth-7c88674d15db6222","modelId":"missingno","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-11","generatedAt":"2026-10-11T20:25:52.518Z","media":{"type":"image","src":"/benchmarks/media/cd090612f6e07a32.png","alt":"MISSINGNO.: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-11T20:25:52.518Z","quality":{"score":42.8,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-11","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is partly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are poorly resolved."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"missingno-svg-raccoon-7e17566f15394e20","modelId":"missingno","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-11","generatedAt":"2026-10-11T20:26:05.444Z","media":{"type":"image","src":"/benchmarks/media/9ff6682e698ab84b.png","alt":"MISSINGNO.: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-11T20:26:05.444Z","quality":{"score":57.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"The requested interaction reads clearly through recognizable animals, expressive faces and a coherent hull-water relationship. Clean curves and readable accessories give the cartoon polish. Partly unresolved grips and supporting contacts limit physical construction; modest shape complexity, weak directional lighting and schematic water constrain its finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"missingno-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"missingno","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-11","generatedAt":"2026-10-11T20:25:51.783Z","media":{"type":"image","src":"/benchmarks/media/9fefde00a39e121f.png","alt":"MISSINGNO.: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-11T20:25:51.783Z","quality":{"score":27.8,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The centered sphere, separate airborne black blob and cyan-above-crimson ordering provide limited valid structure. The colored beads do not establish the required impact and transport histories. One dim floor image supports a small part of the optical system, while the defining wall repetitions, nested sphere imagery and emitter mappings are missing. Rounded volume and some gloss earn material credit, tempered by soft surface appearance and excessive paint-bead scale.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":19,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":7,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":12,"max":120},{"id":"materialsScale","label":"Materials and scale","score":5.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0008856682053828746,"y":0.005024676135349462,"scale":0.8607079764250575},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"missingno-svg-selfie-cb65cc02e50e07af","modelId":"missingno","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A glossy purple jelly struggles with an oversized fake moustache at a playful checkpoint. The scene tests comic expression, physical contact and the appearance of a soft translucent body.","promptHash":"cb65cc02e50e07af","date":"2026-10-11","generatedAt":"2026-10-11T20:47:28.771Z","media":{"type":"image","src":"/benchmarks/media/0694d395667d75c8.png","alt":"MISSINGNO.: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-11T20:47:28.771Z","quality":{"score":49.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-11","rationale":"An appealing purple character with a readable pleased expression and raised-moustache joke. Smooth shading and clean contours support the simple design, but indistinct mitten grips, minimally textured curls and an uncompressed strap weaken the requested physical interaction. The jelly suggests gloss more convincingly than translucency.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A glossy purple jelly struggles with an oversized fake moustache at a playful checkpoint. The scene tests comic expression, physical contact and the appearance of a soft translucent body.","promptSummary":"A glossy purple jelly struggles with an oversized fake moustache at a playful checkpoint. The scene tests comic expression, physical contact and the appearance of a soft translucent body.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-wonder-svg-selfie-462fe51e20c7a75c","modelId":"meta/muse-wonder","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small elephant protecting a delicate seed tests expressive anatomy, physical contact and contrasting organic materials.","promptHash":"462fe51e20c7a75c","date":"2026-10-12","generatedAt":"2026-10-12T20:13:44.618Z","media":{"type":"image","src":"/benchmarks/media/e84437e87c3a833c.png","alt":"Muse Wonder: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-12T20:09:00.442Z","quality":{"score":71.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"An appealing, coherent cartoon with a readable effortful expression and clear forefeet gripping the trunk. Soft colours and controlled contours support its tenderness. Smooth ears, stiff foreleg construction, limited surface texture and uncertain seed contact keep the execution below intricate illustration quality.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small elephant protecting a delicate seed tests expressive anatomy, physical contact and contrasting organic materials.","promptSummary":"A small elephant protecting a delicate seed tests expressive anatomy, physical contact and contrasting organic materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-wonder-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"meta/muse-wonder","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-12","generatedAt":"2026-10-12T20:13:44.618Z","media":{"type":"image","src":"/benchmarks/media/2f8a4cfe53abae58.png","alt":"Muse Wonder: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-12T20:09:00.442Z","quality":{"score":39.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The flattened cyan cap, separate airborne black blob and crimson droplets support much of the contact sequence, although crimson adhesion and forward carriage remain unresolved. Main-sphere placement is broadly compatible with the prescribed view. Sparse isolated mirror images replace the defining repeated room structure and curved nested reflections. The crimson mass is visibly oversized, and the top-center lamp image is missing.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":22,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":12,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":21.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.0000546428823238953,"y":0.008464358739335855,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-wonder-svg-raccoon-7e17566f15394e20","modelId":"meta/muse-wonder","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-12","generatedAt":"2026-10-12T20:13:44.618Z","media":{"type":"image","src":"/benchmarks/media/b27fac9159539c89.png","alt":"Muse Wonder: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-12T20:09:00.442Z","quality":{"score":66.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"An expressive, coherent cartoon demonstrates the requested riding and steering interaction. Clean contours, differentiated accessories and energetic spray support its appeal. Concealed seated contacts, compressed vehicle perspective and weak directional lighting limit physical resolution despite the detailed presentation.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-wonder-svg-pelican-f998b992640eff5e","modelId":"meta/muse-wonder","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-10-12","generatedAt":"2026-10-12T20:13:44.618Z","media":{"type":"image","src":"/benchmarks/media/3c1252f1a5526094.png","alt":"Muse Wonder: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-12T20:09:00.442Z","quality":{"score":72.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"A controlled, expressive seaside illustration with a distinctive pelican, coherent bicycle geometry and successfully resolved graphic detail. Visible pedal support and plausible occlusion strengthen the riding construction. Aligned wheels and layered scenery establish placement. Limited cast shadows, simplified leg articulation and restrained material modeling constrain its physical finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-wonder-svg-mona-lisa-63163c6b35a3998f","modelId":"meta/muse-wonder","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-10-12","generatedAt":"2026-10-12T20:13:44.618Z","media":{"type":"image","src":"/benchmarks/media/bb99a2bf4a11fe6c.png","alt":"Muse Wonder: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-12T20:09:00.442Z","quality":{"score":55.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-12","rationale":"A competent illustrative interpretation with controlled curves, readable details and coherent depth layers. Facial, hair and garment shading demonstrate more development than basic construction. Replication remains limited by the frontal pose, elongated face and pronounced smile; crowded fingers, weak contact shadows and smooth materials constrain anatomical and painterly conviction.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05080527180175231,"y":0.12270436038645613,"scale":1.7419409847740743},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-wonder-svg-ground-truth-7c88674d15db6222","modelId":"meta/muse-wonder","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-12","generatedAt":"2026-10-12T23:53:40.213Z","media":{"type":"image","src":"/benchmarks/media/b9b11551aa039718.png","alt":"Muse Wonder: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-12T20:09:00.442Z","quality":{"score":59.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-18","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-wonder-svg-beach-48eb1c58439ea576","modelId":"meta/muse-wonder","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-10-12","generatedAt":"2026-10-12T20:13:44.618Z","media":{"type":"image","src":"/benchmarks/media/e468bd588fb439c4.png","alt":"Muse Wonder: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-12T20:09:00.442Z","quality":{"score":71,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-20","rationale":"A lively beach illustration with expressive animals, controlled colour transitions and unusually readable accessories. Connected graphic forms make the activities recognizable. Crowded overlaps, compressed depth, schematic joints, weak directional shadows and limited material variation constrain its physical coherence and refinement.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":9},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-4-argon-svg-raccoon-b4cedf42a4a26a4b","modelId":"google/gemini-4-argon","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-13","generatedAt":"2026-10-13T21:10:03.501Z","media":{"type":"image","src":"/benchmarks/media/f5af6a339a32556b.png","alt":"Gemini 4 Argon: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-13T21:10:03.501Z","quality":{"score":80.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"An accomplished illustration with expressive characters, intricate spray and strong depth. Direct back seating, steering connections and hull-water contact are only partly demonstrated. Clean graphic execution and purposeful detail remain strengths, while shadows and reflections provide limited physical grounding.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":8},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-4-argon-svg-pelican-b618cda0f539e818","modelId":"google/gemini-4-argon","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-13","generatedAt":"2026-10-13T21:10:03.501Z","media":{"type":"image","src":"/benchmarks/media/6491a0f52a0b792c.png","alt":"Gemini 4 Argon: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-13T21:10:03.501Z","quality":{"score":84.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"An expressive, detailed illustration with controlled gradients and coherent environmental depth. Continuous body contours and connected bicycle components establish competent construction. Steering grip and pedal engagement remain less explicit than the decorative details and wheel placement. These contact uncertainties limit action and placement credit without establishing broken anatomy.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":8.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":8},{"id":"detail","label":"Purposeful detail","score":8.5},{"id":"artistry","label":"Artistry & character","score":8.5},{"id":"effects","label":"Materials & effects","score":8.5},{"id":"integrity","label":"Visual integrity","score":8.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-4-argon-svg-mona-lisa-58920dfd672bf618","modelId":"google/gemini-4-argon","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-10-13","generatedAt":"2026-10-13T21:10:03.501Z","media":{"type":"image","src":"/benchmarks/media/6999b3bc03e117a6.png","alt":"Gemini 4 Argon: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-13T21:10:03.501Z","quality":{"score":65.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"A coherent, moderately complex Mona Lisa reconstruction with substantial portrait volume, articulated clothing and landscape recession. Specific source features support replication beyond recognition alone. Altered facial likeness, uncertain crossed-hand anatomy and pervasive colored crackle keep the result uneven despite its developed shading and detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.01556205670085376,"y":0.1156251125513383,"scale":1.6160744021928912},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-4-argon-svg-beach-d030aa6f45fe19e9","modelId":"google/gemini-4-argon","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-13","generatedAt":"2026-10-13T21:10:03.501Z","media":{"type":"image","src":"/benchmarks/media/0eccce0c7585a8c1.png","alt":"Gemini 4 Argon: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-13T21:10:03.501Z","quality":{"score":82.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"A strong decorative beach illustration with controlled color, intricate silhouettes and abundant playful detail. Connected dolphin forms and recognizable equipment support its construction, while the aligned sunset reflection anchors the scene. Compressed wave depth, ambiguous foreground support and uneven reflection specificity limit physical coherence. The saved comparisons support retaining the category assessments.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8.5},{"id":"light","label":"Light & reflections","score":8.5},{"id":"detail","label":"Purposeful detail","score":9},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":8},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-4-argon-svg-selfie-bb2b2e8bb4f739a2","modelId":"google/gemini-4-argon","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A luminous little dragon struggles with its bedtime slippers while a buoyant tail pulls it upward. The scene tests expressive anatomy, physical contact and the contrast between translucent forms and heavy objects.","promptHash":"bb2b2e8bb4f739a2","date":"2026-10-13","generatedAt":"2026-10-13T21:10:03.501Z","media":{"type":"image","src":"/benchmarks/media/7d27734df73b83d9.png","alt":"Gemini 4 Argon: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-13T21:10:03.501Z","quality":{"score":71.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-13","rationale":"An appealing, controlled illustration with a distinctive buoyant silhouette, expressive comic face and rich luminous colour. The full composition and grounded slipper read clearly. The principal limitation is the crowded raised-foot interaction: paws, toes and lining overlap without making the bent leg or tight fit unmistakable.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":7.5},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A luminous little dragon struggles with its bedtime slippers while a buoyant tail pulls it upward. The scene tests expressive anatomy, physical contact and the contrast between translucent forms and heavy objects.","promptSummary":"A luminous little dragon struggles with its bedtime slippers while a buoyant tail pulls it upward. The scene tests expressive anatomy, physical contact and the contrast between translucent forms and heavy objects.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-4-argon-svg-ground-truth-7c88674d15db6222","modelId":"google/gemini-4-argon","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-13","generatedAt":"2026-10-13T21:10:03.501Z","media":{"type":"image","src":"/benchmarks/media/17383567cb3403a7.png","alt":"Gemini 4 Argon: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-13T21:10:03.501Z","quality":{"score":68.4,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-13","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response are partly consistent."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail show uneven control."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"google-gemini-4-argon-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"google/gemini-4-argon","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-13","generatedAt":"2026-10-13T21:49:05.561Z","media":{"type":"image","src":"/benchmarks/media/3f00da84f1af34a5.png","alt":"Gemini 4 Argon: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-13T21:49:05.561Z","quality":{"score":58,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-13","rationale":"The central sphere, separate airborne black blob, flattened cyan contact and displaced crimson aftermath support the principal camera and physical-state requirements. Open lateral lanes and initial reflection placements provide partial optical structure, but missing outer chains, sparse nested imagery and incomplete emitter mappings remain substantial limitations. The raised crimson mass also weakens paint-scale fidelity.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":24.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":14.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":50.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":8.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"cat-jam-svg-pelican-b618cda0f539e818","modelId":"cat-jam","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-15","generatedAt":"2026-10-15T03:31:04.143Z","media":{"type":"image","src":"/benchmarks/media/2d352a183e162f39.png","alt":"Cat Jam Free: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-15T03:31:04.143Z","quality":{"score":58,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"A coherent cartoon with readable bicycle placement, controlled outlines and modest surface shading. The separated-looking upper bill weakens pelican likeness; schematic limbs, uncertain gripping contact and simplified mechanics constrain construction. Opaque wheel interiors flatten spatial relationships. Limited feather complexity and directional lighting keep the finish modest.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"cat-jam-svg-selfie-277c5c9d2e092f12","modelId":"cat-jam","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An ivory cat nods intently to music beside a tiny speaker. The scene tests feline anatomy, movement conveyed through a still pose and the contrast between pale fur and woven fabric.","promptHash":"277c5c9d2e092f12","date":"2026-10-15","generatedAt":"2026-10-15T03:31:04.143Z","media":{"type":"image","src":"/benchmarks/media/f56272dd82d70e7c.png","alt":"Cat Jam Free: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-15T03:31:04.143Z","quality":{"score":54.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"A coherent, restrained cat illustration captures the pale palette and intent eyes. Its central expressive requirement—the committed forward-and-down neck bob—is insufficiently articulated. Simple geometry, faint paw construction and minimal surface detail limit the physical presence and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An ivory cat nods intently to music beside a tiny speaker. The scene tests feline anatomy, movement conveyed through a still pose and the contrast between pale fur and woven fabric.","promptSummary":"An ivory cat nods intently to music beside a tiny speaker. The scene tests feline anatomy, movement conveyed through a still pose and the contrast between pale fur and woven fabric.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"cat-jam-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"cat-jam","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-15","generatedAt":"2026-10-15T03:31:04.143Z","media":{"type":"image","src":"/benchmarks/media/890efba1cb507ea6.png","alt":"Cat Jam Free: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-15T03:31:04.143Z","quality":{"score":17.2,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The image conveys a supported central sphere, separate black paint and crimson below cyan. Its contact and transport story is partially resolved. The exterior cutaway camera and largely missing mirror system dominate the limitations; isolated discs and smooth highlights provide little coherent optical structure.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":14,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":4,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":3.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":4,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.0008501843821102561,"y":0.07181159716363873,"scale":0.5915553643668149},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"cat-jam-svg-ground-truth-7c88674d15db6222","modelId":"cat-jam","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-15","generatedAt":"2026-10-15T03:34:46.289Z","media":{"type":"image","src":"/benchmarks/media/cca135b8b0c7c5ab.png","alt":"Cat Jam Free: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-15T03:34:46.289Z","quality":{"score":30.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-16","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"cat-jam-svg-beach-d030aa6f45fe19e9","modelId":"cat-jam","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-15","generatedAt":"2026-10-15T03:37:34.839Z","media":{"type":"image","src":"/benchmarks/media/537e34d45fe1c243.png","alt":"Cat Jam Free: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-15T03:37:34.839Z","quality":{"score":48.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"The scene clearly supplies the requested beach subjects and playful activity. Clean silhouettes and warm gradients support readability, but elementary drawing, ambiguous parasol construction, weak grounding and a rigid sun reflection limit its depth and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"cat-jam-svg-mona-lisa-0b0cd7146594a892","modelId":"cat-jam","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"0b0cd7146594a892","date":"2026-10-15","generatedAt":"2026-10-15T03:38:12.867Z","media":{"type":"image","src":"/benchmarks/media/808b783aa6ef67b3.png","alt":"Cat Jam Free: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-15T03:38:12.867Z","quality":{"score":30.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"The portrait is recognizable and cleanly assembled, but replication remains basic. Its frontal symmetry, elongated silhouette, simplified facial likeness and unresolved hands differ markedly from the original. Modest skin gradients provide volume, while clothing and landscape rely on broad, flat shapes.","categories":[{"id":"fidelity","label":"Task & likeness","score":2.5},{"id":"structure","label":"Anatomy & construction","score":2},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":3.5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":2.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.051336151724011125,"y":0.00114336156847223,"scale":1.6652911949458855},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"cat-jam-svg-raccoon-b4cedf42a4a26a4b","modelId":"cat-jam","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-15","generatedAt":"2026-10-15T03:37:21.759Z","media":{"type":"image","src":"/benchmarks/media/187efc894cf87111.png","alt":"Cat Jam Free: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-15T03:37:21.759Z","quality":{"score":57.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"An appealing cartoon demonstrates the back-riding arrangement through expressive faces and clean shapes. Uncertain steering contact, indistinct deck support and weak hull-water interaction limit physical coherence. Smooth gradients add volume, but simple rounded construction and sparse material detail constrain complexity and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":6},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-open-svg-pelican-f998b992640eff5e","modelId":"meta/muse-spark-open","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-10-15","generatedAt":"2026-10-15T22:01:38.795Z","media":{"type":"image","src":"/benchmarks/media/c5e43ffd96c85ebd.png","alt":"Muse Spark Open: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-15T22:01:38.795Z","quality":{"score":60.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"A controlled, appealing cartoon with a recognizable pelican, expressive face and clean curved forms. Feather lines, spokes and bill shading provide useful complexity. Aligned wheels and a ground shadow establish placement, but uncertain pedal contacts and obscured mechanical connections weaken support and riding action. Lighting remains loosely coordinated with the sun.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-open-svg-selfie-462fe51e20c7a75c","modelId":"meta/muse-spark-open","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small elephant protecting a delicate seed tests expressive anatomy, physical contact and contrasting organic materials.","promptHash":"462fe51e20c7a75c","date":"2026-10-15","generatedAt":"2026-10-15T22:01:38.794Z","media":{"type":"image","src":"/benchmarks/media/24d697f87742df7f.png","alt":"Muse Spark Open: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-15T22:01:38.794Z","quality":{"score":52.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"The rounded silhouette and asymmetric expression make an appealing baby elephant, supported by a quiet background and grounding shadow. The protective action remains difficult to read: forefeet cannot be confidently traced into a grip, and seed contact is crowded by fine lines. Controlled simple shapes provide charm, but limited anatomical clarity, shading and surface detail leave the execution uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small elephant protecting a delicate seed tests expressive anatomy, physical contact and contrasting organic materials.","promptSummary":"A small elephant protecting a delicate seed tests expressive anatomy, physical contact and contrasting organic materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-open-svg-mona-lisa-63163c6b35a3998f","modelId":"meta/muse-spark-open","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-10-15","generatedAt":"2026-10-15T22:01:38.795Z","media":{"type":"image","src":"/benchmarks/media/cbe357a02e308497.png","alt":"Muse Spark Open: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-15T22:01:38.795Z","quality":{"score":37.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"A recognizable, orderly interpretation with clean major contours, connected body forms and modest facial shading. Wide eyes, frontal symmetry and schematic hands limit likeness and expression. Layered scenery establishes basic depth, but shallow arm overlaps, flat clothing and sparse fine detail keep the execution basic and uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":3},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.06655142876885436,"y":0.042907781043346804,"scale":1.4993025000567664},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-open-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"meta/muse-spark-open","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-15","generatedAt":"2026-10-15T22:01:38.795Z","media":{"type":"image","src":"/benchmarks/media/d9aef4570ce92c20.png","alt":"Muse Spark Open: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-15T22:01:38.795Z","quality":{"score":19.6,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The airborne black blob and deformed paint support the broad contact sequence, although cyan attachment and wheel-directed transport remain partial. The sphere and black blob have substantial framing errors. The dominant limitation is the largely absent mirror system: faint wall circles and isolated highlights provide little evidence of recursive room geometry or curved nested reflections. Oversized compact paint features and diffuse material treatment further limit credit.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":14,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":6,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":8,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0.002465821444985531,"y":0.11804813759033472,"scale":0.7408182206817179},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-open-svg-ground-truth-7c88674d15db6222","modelId":"meta/muse-spark-open","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-15","generatedAt":"2026-10-15T22:01:38.795Z","media":{"type":"image","src":"/benchmarks/media/361463ed056fa9b2.png","alt":"Muse Spark Open: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-15T22:01:38.795Z","quality":{"score":51.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-16","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-open-svg-beach-48eb1c58439ea576","modelId":"meta/muse-spark-open","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-10-15","generatedAt":"2026-10-15T22:01:38.795Z","media":{"type":"image","src":"/benchmarks/media/38a38d733b74e17e.png","alt":"Muse Spark Open: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-15T22:01:38.795Z","quality":{"score":64.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"A coherent sunset beach with readable activities, balanced framing and controlled colour transitions. Small accessories add interest, but schematic anatomy, limited surface detail, weak directional shadows and a geometric reflection constrain its finish. Most shapes and overlaps remain clean.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"meta-muse-spark-open-svg-raccoon-7e17566f15394e20","modelId":"meta/muse-spark-open","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-15","generatedAt":"2026-10-15T22:01:38.795Z","media":{"type":"image","src":"/benchmarks/media/2dd6f7c3543447e2.png","alt":"Muse Spark Open: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-15T22:01:38.795Z","quality":{"score":48.1,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-16","rationale":"Expressive faces, distinct accessories and clean color separation make the scene appealing and recognizable. Character detail is stronger than the functional construction: steering contact, limb overlaps and seating remain ambiguous. The readable hull and lively spray establish motion, while limited directional shading and simplified surfaces constrain depth and physical coherence.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":4},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"reflection-beam-1-0-501b-a23b-svg-selfie-a7e8f83efe4ce323","modelId":"reflection/beam-1.0-501b-a23b","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A golden pangolin polishes its oversized amber tail while its nose smudges the same scale. The scene tests expressive anatomy, physical contact and contrasting translucent, polished and soft materials.","promptHash":"a7e8f83efe4ce323","date":"2026-10-16","generatedAt":"2026-10-16T17:27:01.630Z","media":{"type":"image","src":"/benchmarks/media/261c12543ea7cc7c.png","alt":"Beam 1.0 501B A23B: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-16T17:27:01.630Z","quality":{"score":45.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"An appealing expression and warm palette support a coherent simple cartoon. The central polishing joke is not legible: the cloth is unresolved and the nose sits away from the fan. Pointed radial plates, uncertain limb attachments and weak ground contact limit construction, while restrained surface detail limits material richness.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A golden pangolin polishes its oversized amber tail while its nose smudges the same scale. The scene tests expressive anatomy, physical contact and contrasting translucent, polished and soft materials.","promptSummary":"A golden pangolin polishes its oversized amber tail while its nose smudges the same scale. The scene tests expressive anatomy, physical contact and contrasting translucent, polished and soft materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"reflection-beam-1-0-501b-a23b-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"reflection/beam-1.0-501b-a23b","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-16","generatedAt":"2026-10-16T17:27:01.631Z","media":{"type":"image","src":"/benchmarks/media/6070f497f427e904.png","alt":"Beam 1.0 501B A23B: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-16T17:27:01.631Z","quality":{"score":21.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Central object placement is fairly close, but the defining mirrored-room structure is missing. The paint colors suggest the required contact order without establishing their different interaction histories. The faint floor oval does not demonstrate valid reflection geometry, so it earns no reflection credit. Rounded volume and glossy paint provide limited material credit despite weak metallic finishes and oversized paint bodies.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":20.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":11.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":2,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"reflection-beam-1-0-501b-a23b-svg-mona-lisa-0b0cd7146594a892","modelId":"reflection/beam-1.0-501b-a23b","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"0b0cd7146594a892","date":"2026-10-16","generatedAt":"2026-10-16T17:30:37.454Z","media":{"type":"image","src":"/benchmarks/media/2f8968db6173e580.png","alt":"Beam 1.0 501B A23B: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-16T17:30:37.454Z","quality":{"score":0.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"The candidate shows a faint warm oval against an almost entirely black field. Minimal placement and smooth tonal variation are visible, but the portrait, anatomy, crossed hands and landscape remain unreadable. The saved comparisons provide no specific evidence supporting a change to the baseline categories.","categories":[{"id":"fidelity","label":"Task & likeness","score":0},{"id":"structure","label":"Anatomy & construction","score":0},{"id":"space","label":"Space & placement","score":0},{"id":"light","label":"Light & reflections","score":0.5},{"id":"detail","label":"Purposeful detail","score":0},{"id":"artistry","label":"Artistry & character","score":0},{"id":"effects","label":"Materials & effects","score":0},{"id":"integrity","label":"Visual integrity","score":0}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"reflection-beam-1-0-501b-a23b-svg-pelican-ae766dae25606b2d","modelId":"reflection/beam-1.0-501b-a23b","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"ae766dae25606b2d","date":"2026-10-16","generatedAt":"2026-10-16T17:30:56.422Z","media":{"type":"image","src":"/benchmarks/media/57083c3dcf36bb3a.png","alt":"Beam 1.0 501B A23B: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-16T17:30:56.422Z","quality":{"score":46.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"A cheerful, clearly readable pelican-and-bicycle illustration with clean contours and balanced composition. Simple geometric construction limits artistic complexity, while ambiguous neck and steering overlaps, merged pedal contacts and an incomplete drivetrain weaken physical coherence.","categories":[{"id":"fidelity","label":"Task & likeness","score":5.5},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"reflection-beam-1-0-501b-a23b-svg-raccoon-ffbfac3e47be1fd2","modelId":"reflection/beam-1.0-501b-a23b","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"ffbfac3e47be1fd2","date":"2026-10-16","generatedAt":"2026-10-16T17:31:24.076Z","media":{"type":"image","src":"/benchmarks/media/41ac6a68d8c0b30f.png","alt":"Beam 1.0 501B A23B: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-16T17:31:24.076Z","quality":{"score":55.2,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"A cheerful, readable illustration with recognizable animals, expressive faces and useful accessories. Simple rounded construction limits complexity, while ambiguous steering attachments and weak hull-to-water contact reduce physical coherence. Lighting and surface effects establish the setting without convincing environmental interaction.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"reflection-beam-1-0-501b-a23b-svg-beach-624714eb8b19d3ee","modelId":"reflection/beam-1.0-501b-a23b","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"624714eb8b19d3ee","date":"2026-10-16","generatedAt":"2026-10-16T17:31:49.433Z","media":{"type":"image","src":"/benchmarks/media/6dea244dd66e5ac6.png","alt":"Beam 1.0 501B A23B: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-16T17:31:49.433Z","quality":{"score":52.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"A coherent, cheerful beach illustration fulfills the main subjects with clean shapes and a strong sunset focal point. Its basic geometric construction, ambiguous animal anatomy, shallow depth and schematic lighting limit its accomplishment. Decorative details provide interest without developing intricate forms or convincing physical surfaces.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"reflection-beam-1-0-501b-a23b-svg-ground-truth-d3064be697f0e4e2","modelId":"reflection/beam-1.0-501b-a23b","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"d3064be697f0e4e2","date":"2026-10-16","generatedAt":"2026-10-16T17:32:10.729Z","media":{"type":"image","src":"/benchmarks/media/70b53759fe67074f.png","alt":"Beam 1.0 501B A23B: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-16T17:32:10.729Z","quality":{"score":27.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-18","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are partly correct."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-1017-svg-beach-d030aa6f45fe19e9","modelId":"deepseek/deepseek-v4.1-flash-1017","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-18","generatedAt":"2026-10-18T17:26:58.850Z","media":{"type":"image","src":"/benchmarks/media/ddf2f6caa8b17429.png","alt":"DeepSeek-V4.1-Flash-1017: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-18T17:26:58.850Z","quality":{"score":57.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-22","rationale":"A coherent, colorful beach illustration with controlled curves, clean layering and varied readable accessories. Castle details and small animal features strengthen its finish. Schematic airborne animals, uncertain board support, shallow architectural depth and decorative reflections limit physical coherence and drawing sophistication.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":4.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-1017-svg-mona-lisa-e7b85d8cd393c26f","modelId":"deepseek/deepseek-v4.1-flash-1017","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-18","generatedAt":"2026-10-18T17:26:58.849Z","media":{"type":"image","src":"/benchmarks/media/381887c96fb29aa4.png","alt":"DeepSeek-V4.1-Flash-1017: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-18T17:26:58.849Z","quality":{"score":54.6,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"A coherent graphic interpretation with controlled contours, modeled facial volume and purposeful portrait details. Replication remains limited by the frontal pose, enlarged face, altered expression and ambiguous crossed hands. Landscape layering provides depth, while schematic fabric, smooth surfaces and tonal bands constrain the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05299400050246094,"y":0.1678088616399928,"scale":1.8221188003905084},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-1017-svg-pelican-b618cda0f539e818","modelId":"deepseek/deepseek-v4.1-flash-1017","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-18","generatedAt":"2026-10-18T17:26:59.407Z","media":{"type":"image","src":"/benchmarks/media/b9ecce7a10efed93.png","alt":"DeepSeek-V4.1-Flash-1017: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-18T17:26:59.407Z","quality":{"score":73.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"An expressive, controlled illustration with continuous body construction, substantial bicycle detail and effective landscape depth. Unclear foot-to-pedal and wing-to-handlebar contacts limit physical conviction and local readability. Purposeful detail and composition are stronger than directional lighting and material treatment.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-1017-svg-ground-truth-7c88674d15db6222","modelId":"deepseek/deepseek-v4.1-flash-1017","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-18","generatedAt":"2026-10-18T17:26:59.468Z","media":{"type":"image","src":"/benchmarks/media/7d284e3341f8dac6.png","alt":"DeepSeek-V4.1-Flash-1017: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-18T17:26:59.468Z","quality":{"score":60.6,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-18","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is mostly consistent."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are clean and well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-1017-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"deepseek/deepseek-v4.1-flash-1017","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-18","generatedAt":"2026-10-18T17:26:59.468Z","media":{"type":"image","src":"/benchmarks/media/1a94233e3e67f747.png","alt":"DeepSeek-V4.1-Flash-1017: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-18T17:26:59.468Z","quality":{"score":48.7,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"Principal sphere and airborne-black placements are supported, and diminishing bilateral chains provide limited optical structure. Missing nested imagery, incomplete outer chains, distorted first images and sparse emitter recurrence remain the dominant weaknesses. Oversized paint bodies and unresolved cyan contact impose separate limitations.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":13.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":34,"max":120},{"id":"materialsScale","label":"Materials and scale","score":8.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-1017-svg-selfie-c14ada77a4644cf8","modelId":"deepseek/deepseek-v4.1-flash-1017","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptHash":"c14ada77a4644cf8","date":"2026-10-18","generatedAt":"2026-10-18T19:07:59.783Z","media":{"type":"image","src":"/benchmarks/media/fa237b30bbec4f98.png","alt":"DeepSeek-V4.1-Flash-1017: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-18T19:07:59.783Z","quality":{"score":74.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"An expressive, intricately drawn storybook scene with a complete flowing silhouette and differentiated glossy materials. The supported main lens and continuous shrimp body make the central interaction readable, but reciprocal eye lines, facial magnification and the smaller lens grip remain ambiguous. Bright granular fringes and crowded accessory lines limit the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":8},{"id":"effects","label":"Materials & effects","score":7.5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptSummary":"A whale and a tiny shrimp examine one another through bubble lenses, with an octopus nearby. The scene tests interacting characters, differences in scale and transparent materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"deepseek-deepseek-v4-1-flash-1017-svg-raccoon-17d860fae63dc1c6","modelId":"deepseek/deepseek-v4.1-flash-1017","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"17d860fae63dc1c6","date":"2026-10-18","generatedAt":"2026-10-18T21:26:03.765Z","media":{"type":"image","src":"/benchmarks/media/f7f40b031db33819.png","alt":"DeepSeek-V4.1-Flash-1017: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-18T21:26:03.765Z","quality":{"score":71.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-18","rationale":"An appealing cartoon with recognizable riders, expressive faces and a cohesive twilight palette. Clean shapes support the composition, but ambiguous limb overlaps, weak seat contact and loosely resolved steering limit physical coherence. Lighting and water effects are decorative, with modest material differentiation and drawing complexity.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-1-flash-svg-pelican-f998b992640eff5e","modelId":"minimax/minimax-m3.1-flash","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"f998b992640eff5e","date":"2026-10-19","generatedAt":"2026-10-19T22:49:02.956Z","media":{"type":"image","src":"/benchmarks/media/07a2e448b4497629.png","alt":"MiniMax M3.1 Flash: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-19T22:49:02.956Z","quality":{"score":72,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-20","rationale":"A controlled coastal illustration with a clear pelican silhouette, detailed wheels and accomplished feather and bill drawing. Continuous body construction and grounded wheels support the scene. Ambiguous leg, pedal and steering contacts weaken the riding mechanics; approximate directional lighting and simplified materials limit physical conviction.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":6.5},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":7},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-1-flash-svg-selfie-d50ad6124d137ee4","modelId":"minimax/minimax-m3.1-flash","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small robot with an oversized angular head presents a glowing object. The scene tests character proportions, gesture and contrasting materials.","promptHash":"d50ad6124d137ee4","date":"2026-10-19","generatedAt":"2026-10-19T22:49:02.953Z","media":{"type":"image","src":"/benchmarks/media/2259c1ebc40b035c.png","alt":"MiniMax M3.1 Flash: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-19T22:49:02.953Z","quality":{"score":61.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-20","rationale":"The sculptural M, confident grin and miniature theatrical pose closely fulfill the brief. Controlled gradients and clean shapes give the character volume and appeal. Moderate drawing complexity, an apparently separated shoe and ambiguous grounding limit the finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":7.5},{"id":"structure","label":"Anatomy & construction","score":6},{"id":"space","label":"Space & placement","score":7},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small robot with an oversized angular head presents a glowing object. The scene tests character proportions, gesture and contrasting materials.","promptSummary":"A small robot with an oversized angular head presents a glowing object. The scene tests character proportions, gesture and contrasting materials.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-1-flash-svg-mona-lisa-63163c6b35a3998f","modelId":"minimax/minimax-m3.1-flash","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"63163c6b35a3998f","date":"2026-10-19","generatedAt":"2026-10-19T22:49:02.953Z","media":{"type":"image","src":"/benchmarks/media/064be4e926bf6ca8.png","alt":"MiniMax M3.1 Flash: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-19T22:49:02.953Z","quality":{"score":50,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-20","rationale":"A coherent decorative portrait combines layered hair, garment contours, linked trim and atmospheric scenery. These provide meaningful drawing complexity, but likeness remains limited by the frontal elongated face and altered clothing. Compressed hands, abrupt facial shading and cluttered local edges leave the portrait uneven.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":4.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":5.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04588323226590277,"y":0.042622910864911275,"scale":1.5683121334412315},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-1-flash-svg-raccoon-7e17566f15394e20","modelId":"minimax/minimax-m3.1-flash","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-19","generatedAt":"2026-10-19T22:49:02.956Z","media":{"type":"image","src":"/benchmarks/media/212128418b9af112.png","alt":"MiniMax M3.1 Flash: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-19T22:49:02.956Z","quality":{"score":76.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-20","rationale":"An expressive, coherent cartoon clearly demonstrates the requested riding and steering relationship. Layered animal shapes, readable vehicle depth and purposeful facial details support its drawing quality. Ambiguous grips, sparse cast shadows and simplified material responses limit physical precision.","categories":[{"id":"fidelity","label":"Task & likeness","score":9},{"id":"structure","label":"Anatomy & construction","score":8},{"id":"space","label":"Space & placement","score":8},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":8}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-1-flash-svg-ground-truth-7c88674d15db6222","modelId":"minimax/minimax-m3.1-flash","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-19","generatedAt":"2026-10-19T22:49:02.961Z","media":{"type":"image","src":"/benchmarks/media/f379002a546d1424.png","alt":"MiniMax M3.1 Flash: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-19T22:49:02.961Z","quality":{"score":41.9,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-20","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-1-flash-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"minimax/minimax-m3.1-flash","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-19","generatedAt":"2026-10-19T22:49:02.959Z","media":{"type":"image","src":"/benchmarks/media/ad8015e73b6c9f24.png","alt":"MiniMax M3.1 Flash: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-19T22:49:02.959Z","quality":{"score":39.8,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-20","rationale":"The sphere placement, separate airborne black blob, recent cyan contact and elongated crimson aftermath support the required gross physical state. The reflection system remains severely incomplete: isolated side images, broad metallic bands and sparse light patches replace extensive mirror repetitions and curved nested imagery.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":20,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":10,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":24,"max":120},{"id":"materialsScale","label":"Materials and scale","score":8,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"minimax-minimax-m3-1-flash-svg-beach-48eb1c58439ea576","modelId":"minimax/minimax-m3.1-flash","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"48eb1c58439ea576","date":"2026-10-19","generatedAt":"2026-10-19T22:49:02.959Z","media":{"type":"image","src":"/benchmarks/media/bc06245337843354.png","alt":"MiniMax M3.1 Flash: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-19T22:49:02.959Z","quality":{"score":71.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-20","rationale":"A controlled, detailed beach illustration with expressive animals, coherent depth and accomplished foliage and accessory drawing. Its warm palette and varied contours sustain visual interest. Simplified anatomy and object volumes limit physical conviction, while reflection streaks crossing dry-looking sand and generic shadows weaken the lighting.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7},{"id":"space","label":"Space & placement","score":7.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":8},{"id":"artistry","label":"Artistry & character","score":6.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-pro-svg-mona-lisa-e7b85d8cd393c26f","modelId":"xiaomi/mimo-v2.6-pro","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"e7b85d8cd393c26f","date":"2026-10-22","generatedAt":"2026-10-22T15:29:30.662Z","media":{"type":"image","src":"/benchmarks/media/d36486b582250c9f.png","alt":"MiMo-V2.6-Pro: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-22T15:29:30.662Z","quality":{"score":39.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-22","rationale":"The portrait is recognizable through its dark silhouette, slight smile and muted landscape, but replication is weak. A frontal, oversized head and symmetric torso replace the reference's turned seated pose; angular, separated hands omit its central gesture. Smooth shading maintains coherence while anatomical detail, depth and material rendering remain basic.","categories":[{"id":"fidelity","label":"Task & likeness","score":3.5},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":3},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.05154283130414031,"y":0.18567578391095477,"scale":1.7419409847740746},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-pro-svg-beach-d030aa6f45fe19e9","modelId":"xiaomi/mimo-v2.6-pro","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-22","generatedAt":"2026-10-22T15:29:30.698Z","media":{"type":"image","src":"/benchmarks/media/79e22c63a3e1c1ae.png","alt":"MiMo-V2.6-Pro: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-22T15:29:30.698Z","quality":{"score":60.9,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-22","rationale":"A readable sunset beach with clear requested subjects, appealing framing and purposeful accessories. Schematic anatomy, flattened foreground objects, repetitive water highlights and sparse directional shadows limit construction and physical depth. The kite's unresolved anchor also constrains integrity.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-pro-svg-selfie-271fb2af67f0ddcd","modelId":"xiaomi/mimo-v2.6-pro","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"A small rabbit in an explorer outfit floats with a luminous orb. The scene tests character anatomy, pose and the placement of geometric forms around it.","promptHash":"271fb2af67f0ddcd","date":"2026-10-22","generatedAt":"2026-10-22T15:29:30.663Z","media":{"type":"image","src":"/benchmarks/media/feaf4b603ea80840.png","alt":"MiMo-V2.6-Pro: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-22T15:29:30.663Z","quality":{"score":58,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-22","rationale":"An appealing rabbit with balanced framing, expressive eyes and purposeful costume accents. Simple frontal construction limits anatomical and spatial accomplishment. The orange orb provides an effective luminous focal point, but unreadable wrist connections weaken the requested gestures and prevent stronger integrity credit.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A small rabbit in an explorer outfit floats with a luminous orb. The scene tests character anatomy, pose and the placement of geometric forms around it.","promptSummary":"A small rabbit in an explorer outfit floats with a luminous orb. The scene tests character anatomy, pose and the placement of geometric forms around it.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-pro-svg-pelican-b618cda0f539e818","modelId":"xiaomi/mimo-v2.6-pro","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-22","generatedAt":"2026-10-22T15:29:30.663Z","media":{"type":"image","src":"/benchmarks/media/bf869d8e8843fffa.png","alt":"MiMo-V2.6-Pro: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-22T15:29:30.663Z","quality":{"score":58,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-22","rationale":"The distinctive bill, continuous silhouette and detailed wheels establish a readable pelican cyclist. Shaped feathers and soft shading add character, but jointless legs, crowded pedal contacts, absent steering contact and loosely directional lighting limit the construction. The comparisons support retaining the baseline category scores.","categories":[{"id":"fidelity","label":"Task & likeness","score":6.5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-pro-svg-raccoon-b4cedf42a4a26a4b","modelId":"xiaomi/mimo-v2.6-pro","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-22","generatedAt":"2026-10-22T17:15:53.927Z","media":{"type":"image","src":"/benchmarks/media/ce6f15ce88658b7b.png","alt":"MiMo-V2.6-Pro: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-22T17:15:53.927Z","quality":{"score":56.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-22","rationale":"A coherent, cheerful cartoon with recognizable animals, readable steering contact and useful small details. The dog's orientation follows the craft more successfully than a wholly frontal pose would. Simple anatomy, weak raccoon grip and compressed seating contacts still limit physical credibility. Gradients and spray provide polish without resolving complex construction.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":6},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":5.5},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-pro-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"xiaomi/mimo-v2.6-pro","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-22","generatedAt":"2026-10-22T17:37:36.323Z","media":{"type":"image","src":"/benchmarks/media/32f40157675eaf68.png","alt":"MiMo-V2.6-Pro: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-22T17:37:36.323Z","quality":{"score":39,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-22","rationale":"The main framing and visible paint-contact story are well supported. Major optical omissions dominate the result: sparse repetitions, undersized first images, missing outer chains and absent curved nested imagery prevent a coherent mirror-room system. The metallic forms remain convincing in volume but have a soft illustrative finish.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":12,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":19,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7.5,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"xiaomi-mimo-v2-6-pro-svg-ground-truth-7c88674d15db6222","modelId":"xiaomi/mimo-v2.6-pro","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-22","generatedAt":"2026-10-22T19:03:43.850Z","media":{"type":"image","src":"/benchmarks/media/f49f05ea94037a5d.png","alt":"MiMo-V2.6-Pro: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-22T19:03:43.850Z","quality":{"score":43.6,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-22","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are mostly consistent."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response are partly consistent."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are mostly well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"ox-alpha-svg-raccoon-7e17566f15394e20","modelId":"ox-alpha","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"7e17566f15394e20","date":"2026-10-23","generatedAt":"2026-10-23T00:32:50.219Z","media":{"type":"image","src":"/benchmarks/media/a485cc85138c5002.png","alt":"Ox Alpha: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-23T00:32:50.219Z","quality":{"score":56.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"Recognizable animals, a supported riding pose and an energetic diagonal composition make the scene readable. Character details and spray strengthen its appeal. Simple limb construction, limited directional lighting and an unresolved steering connection constrain the drawing's physical precision and complexity.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":6}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-1-pro-svg-mona-lisa-58920dfd672bf618","modelId":"moonshotai/kimi-k3.1-pro","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"58920dfd672bf618","date":"2026-10-23","generatedAt":"2026-10-23T18:27:48.579Z","media":{"type":"image","src":"/benchmarks/media/6c04ce0adba09028.png","alt":"Kimi K3.1 Pro: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-23T18:27:48.579Z","quality":{"score":55.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"A coherent, recognizable interpretation with controlled color and readable landscape layers. Replication is limited by the frontal seated construction, elongated facial proportions and oversized, weakly articulated hands. Smooth shading supplies some volume but little of the reference's subtle expression, fabric structure or atmospheric transitions.","categories":[{"id":"fidelity","label":"Task & likeness","score":5},{"id":"structure","label":"Anatomy & construction","score":5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":5},{"id":"integrity","label":"Visual integrity","score":6.5}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.040968090008964765,"y":-0.04114948985957939,"scale":1.2904616208728896},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-1-pro-svg-raccoon-b4cedf42a4a26a4b","modelId":"moonshotai/kimi-k3.1-pro","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"b4cedf42a4a26a4b","date":"2026-10-23","generatedAt":"2026-10-23T18:27:48.589Z","media":{"type":"image","src":"/benchmarks/media/ed60d4ecfa6bfa57.png","alt":"Kimi K3.1 Pro: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-23T18:27:48.589Z","quality":{"score":61.8,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The additional comparisons support retaining the assessment. Readable steering, shaped animal contours and purposeful details coexist with simplified joints, a frontal passenger whose seated weight is weakly conveyed, and limited physical lighting. The angled craft establishes more depth than the riding pose resolves.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":5.5},{"id":"detail","label":"Purposeful detail","score":7},{"id":"artistry","label":"Artistry & character","score":6},{"id":"effects","label":"Materials & effects","score":6},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-1-pro-svg-selfie-51df87a90bcca7ff","modelId":"moonshotai/kimi-k3.1-pro","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptHash":"51df87a90bcca7ff","date":"2026-10-23","generatedAt":"2026-10-23T18:27:48.578Z","media":{"type":"image","src":"/benchmarks/media/279946528764ba2f.png","alt":"Kimi K3.1 Pro: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-23T18:27:48.578Z","quality":{"score":67.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"The transformed reflection, attentive pose and constellation anatomy clearly communicate introspection. Purposeful fine details enrich a coherent composition. Tubular limbs, crowded lower-body overlaps, shallow depth and limited cast lighting keep the overall execution within the competent range.","categories":[{"id":"fidelity","label":"Task & likeness","score":8},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7},{"id":"effects","label":"Materials & effects","score":7},{"id":"integrity","label":"Visual integrity","score":7}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummary":"An imagined creature encounters its reflection. The scene tests original character design, visual self-interpretation and the relationship between a subject and its reflected image.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-1-pro-svg-pelican-b618cda0f539e818","modelId":"moonshotai/kimi-k3.1-pro","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"b618cda0f539e818","date":"2026-10-23","generatedAt":"2026-10-23T18:27:48.579Z","media":{"type":"image","src":"/benchmarks/media/7eaf82480a0114f9.png","alt":"Kimi K3.1 Pro: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-23T18:27:48.579Z","quality":{"score":73.5,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"A characterful coastal illustration with controlled contours and varied feather, accessory and bicycle detail. Layered scenery and subtle shading enrich the presentation. Soft leg articulation, unresolved riding contacts, shallow depth and generalized lighting limit physical precision.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":7.5},{"id":"space","label":"Space & placement","score":6.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":7.5},{"id":"artistry","label":"Artistry & character","score":7.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-1-pro-svg-beach-d030aa6f45fe19e9","modelId":"moonshotai/kimi-k3.1-pro","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"d030aa6f45fe19e9","date":"2026-10-23","generatedAt":"2026-10-23T18:27:48.589Z","media":{"type":"image","src":"/benchmarks/media/8c7ed2ebd141391a.png","alt":"Kimi K3.1 Pro: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-23T18:27:48.589Z","quality":{"score":64.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-23","rationale":"A coherent, playful beach illustration with readable animals, an organized composition and an aligned sunset reflection. Connected silhouettes, sand texture and atmospheric effects demonstrate controlled execution. Simplified joints, flattened foreground viewpoints and inconsistent shadows limit anatomical and spatial credibility.","categories":[{"id":"fidelity","label":"Task & likeness","score":8.5},{"id":"structure","label":"Anatomy & construction","score":5.5},{"id":"space","label":"Space & placement","score":5.5},{"id":"light","label":"Light & reflections","score":6.5},{"id":"detail","label":"Purposeful detail","score":6.5},{"id":"artistry","label":"Artistry & character","score":5.5},{"id":"effects","label":"Materials & effects","score":6.5},{"id":"integrity","label":"Visual integrity","score":7.5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-1-pro-svg-ground-truth-7c88674d15db6222","modelId":"moonshotai/kimi-k3.1-pro","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"7c88674d15db6222","date":"2026-10-23","generatedAt":"2026-10-23T18:27:48.599Z","media":{"type":"image","src":"/benchmarks/media/ebd9078fefce3747.png","alt":"Kimi K3.1 Pro: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-23T18:27:48.599Z","quality":{"score":45.3,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-23","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout closely match the scene."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction are mostly well controlled."},{"label":"Colour and finish","text":"Colour and surface finish are convincing and well controlled."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"moonshotai-kimi-k3-1-pro-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"moonshotai/kimi-k3.1-pro","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-23","generatedAt":"2026-10-23T18:27:48.589Z","media":{"type":"image","src":"/benchmarks/media/2d1ba7b8c8b08f44.png","alt":"Kimi K3.1 Pro: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-23T18:27:48.589Z","quality":{"score":41.1,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-23","rationale":"The settled sphere, separate airborne black paint and crimson-before-cyan interaction are supported, with a modest crimson travel overshoot. A shared downward framing displacement remains after accounting for the square canvas. The principal limitation is the reflection system: sparse sphere copies and generic curved bands omit most of the ordered mirror-room structure. Materials receive a small upward revision because bounded bright images and dark bands visibly establish some polish, while soft shading and bulky paint remain separate weaknesses.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":23.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":10.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":26,"max":120},{"id":"materialsScale","label":"Materials and scale","score":7,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-4.336808689942018e-19,"y":0.06871165644171782,"scale":1},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"iceman-contributor-svg-mona-lisa-0b0cd7146594a892","modelId":"iceman-contributor","groupId":"mona-lisa","kind":"svg","title":"Mona Lisa","prompt":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptHash":"0b0cd7146594a892","date":"2026-10-25","generatedAt":"2026-10-25T04:19:55.279Z","media":{"type":"image","src":"/benchmarks/media/064f78e3888cdb1d.png","alt":"Iceman Contributor: Mona Lisa SVG benchmark"},"firstGeneratedAt":"2026-10-25T04:19:55.279Z","quality":{"score":19.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-25","rationale":"The portrait retains a dark silhouette against an earthy landscape, but replication is weak. The frontal body, elongated facial proportions and undifferentiated hands lose the reference's defining pose and likeness. Basic gradients and layered hills provide some coherence, with little anatomical or painterly detail.","categories":[{"id":"fidelity","label":"Task & likeness","score":1.5},{"id":"structure","label":"Anatomy & construction","score":1.5},{"id":"space","label":"Space & placement","score":2.5},{"id":"light","label":"Light & reflections","score":2},{"id":"detail","label":"Purposeful detail","score":2},{"id":"artistry","label":"Artistry & character","score":2},{"id":"effects","label":"Materials & effects","score":2},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":8,"scale":10},"referenceAlignment":{"x":0.04015657430580668,"y":0.4373676798255242,"scale":1.4119899196676586},"taskDescription":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummary":"An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"iceman-contributor-svg-pelican-ae766dae25606b2d","modelId":"iceman-contributor","groupId":"pelican","kind":"svg","title":"Pelican","prompt":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptHash":"ae766dae25606b2d","date":"2026-10-25","generatedAt":"2026-10-25T04:19:55.283Z","media":{"type":"image","src":"/benchmarks/media/2cb3ee72e4bc9d4b.png","alt":"Iceman Contributor: Pelican SVG benchmark"},"firstGeneratedAt":"2026-10-25T04:19:55.283Z","quality":{"score":38,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-25","rationale":"A recognizable pelican and bicycle sit within an appealing sunset composition. Feather contours, bill shape and atmospheric colour provide character, but disconnected-looking leg origins, unclear pedal contact and unresolved steering weaken the central riding action. Consistent wheel alignment establishes a side view without resolving the rider's support.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":3.5},{"id":"light","label":"Light & reflections","score":5},{"id":"detail","label":"Purposeful detail","score":4.5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":0,"visual":3,"scale":10},"taskDescription":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummary":"A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"iceman-contributor-svg-raccoon-ffbfac3e47be1fd2","modelId":"iceman-contributor","groupId":"raccoon","kind":"svg","title":"Raccoon","prompt":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptHash":"ffbfac3e47be1fd2","date":"2026-10-25","generatedAt":"2026-10-25T04:19:55.279Z","media":{"type":"image","src":"/benchmarks/media/46aa3e8277cb34c8.png","alt":"Iceman Contributor: Raccoon SVG benchmark"},"firstGeneratedAt":"2026-10-25T04:19:55.279Z","quality":{"score":36.7,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-25","rationale":"The sunset palette and red hull establish an appealing setting, but the central interaction remains unresolved. Overlapping bodies obscure the dog, supported seating is unclear, and no readable paw-to-control grip demonstrates steering. Simple construction, weak environmental lighting and loosely described spray limit depth and finish.","categories":[{"id":"fidelity","label":"Task & likeness","score":3},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":4.5},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4},{"id":"effects","label":"Materials & effects","score":4},{"id":"integrity","label":"Visual integrity","score":3}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummary":"A jet-ski scene pairs a canine driver with a raccoon passenger. It tests the placement and contact of interacting subjects, along with vehicle construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"iceman-contributor-svg-beach-624714eb8b19d3ee","modelId":"iceman-contributor","groupId":"beach","kind":"svg","title":"Beach","prompt":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptHash":"624714eb8b19d3ee","date":"2026-10-25","generatedAt":"2026-10-25T04:19:55.283Z","media":{"type":"image","src":"/benchmarks/media/8099512fd520b2a2.png","alt":"Iceman Contributor: Beach SVG benchmark"},"firstGeneratedAt":"2026-10-25T04:19:55.283Z","quality":{"score":46.4,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-25","rationale":"A coherent sunset beach with recognizable animals, playful details and effective palm framing. Connected silhouettes and layered scenery support modest construction and placement credit. Jointless figures, limited foreshortening and an inadequately separated sun and reflection keep the execution basic despite its attractive color.","categories":[{"id":"fidelity","label":"Task & likeness","score":7},{"id":"structure","label":"Anatomy & construction","score":3.5},{"id":"space","label":"Space & placement","score":6},{"id":"light","label":"Light & reflections","score":4},{"id":"detail","label":"Purposeful detail","score":5.5},{"id":"artistry","label":"Artistry & character","score":3.5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":5}],"maximum":100},"taskComplexity":{"cognitive":1,"visual":5,"scale":10},"taskDescription":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummary":"A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"iceman-contributor-svg-newtons-mirror-5a00e74f7bfe6e8e","modelId":"iceman-contributor","groupId":"newtons-mirror","kind":"svg","title":"Newton's Mirror","prompt":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptHash":"5a00e74f7bfe6e8e","date":"2026-10-25","generatedAt":"2026-10-25T04:19:55.306Z","media":{"type":"image","src":"/benchmarks/media/695b2a08a00bba93.png","alt":"Iceman Contributor: Newton's Mirror SVG benchmark"},"firstGeneratedAt":"2026-10-25T04:19:55.306Z","quality":{"score":10.1,"rubricId":"newtons-mirror-physics-v5","reviewedAt":"2026-10-25","rationale":"The image provides limited evidence through rounded volume, faint wall strips and floor echoes. Its main sphere is small and low, the paint beads do not establish the required interaction history, and most of the defining mirror structure is missing.","categories":[{"id":"physicalStateTiming","label":"State at 600 ms","score":0.5,"max":25},{"id":"cameraComposition","label":"Camera and placement","score":2.5,"max":15},{"id":"reflectionConsistency","label":"Reflections","score":7.5,"max":120},{"id":"materialsScale","label":"Materials and scale","score":3,"max":10},{"id":"floorDamage","label":"Floor-damage reward","score":0,"max":10}],"maximum":100,"scoreFormat":"normalized-100-v1"},"taskComplexity":{"cognitive":7,"visual":7,"scale":10},"referenceAlignment":{"x":-0.0020772597828205386,"y":0.07608416244513222,"scale":0.6872892787909721},"taskDescription":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummary":"A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"iceman-contributor-svg-ground-truth-d3064be697f0e4e2","modelId":"iceman-contributor","groupId":"ground-truth","kind":"svg","title":"Ground Truth","prompt":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptHash":"d3064be697f0e4e2","date":"2026-10-25","generatedAt":"2026-10-25T04:19:55.307Z","media":{"type":"image","src":"/benchmarks/media/2790bb6143c194f8.png","alt":"Iceman Contributor: Ground Truth SVG benchmark"},"firstGeneratedAt":"2026-10-25T04:19:55.307Z","quality":{"score":20.2,"maximum":100,"rubricId":"ground-truth-dual-v7","reviewedAt":"2026-10-25","scoreFormat":"independent-capabilities-v1","capabilityFindings":{"general":[{"label":"Camera and layout","text":"Camera and layout are poorly resolved."},{"label":"Physical state","text":"The physical state is not clearly resolved."},{"label":"Lighting and surfaces","text":"Lighting and surface response have major weaknesses."}],"visual":[{"label":"Shapes and construction","text":"Shapes and construction show uneven control."},{"label":"Colour and finish","text":"Colour and surface finish show uneven control."},{"label":"Light and shadow detail","text":"Light and shadow detail are poorly resolved."}]},"categories":[]},"taskComplexity":{"cognitive":8,"visual":5,"scale":10},"referenceAlignment":{"x":0,"y":0,"scale":1},"taskDescription":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummary":"A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1},{"id":"iceman-contributor-svg-selfie-d2cbace4cde7f6b9","modelId":"iceman-contributor","groupId":"selfie","kind":"svg","title":"Selfie","prompt":"Character scenes test expression, anatomy, materials and coherent visual construction.","promptHash":"d2cbace4cde7f6b9","date":"2026-10-25","generatedAt":"2026-10-25T04:40:13.391Z","media":{"type":"image","src":"/benchmarks/media/985df66258e298d8.png","alt":"Iceman Contributor: Selfie SVG benchmark"},"firstGeneratedAt":"2026-10-25T04:40:13.391Z","quality":{"score":39.3,"rubricId":"svg-visual-v2","reviewedAt":"2026-10-25","rationale":"The crystal crown, cyan eyes and blue faceting communicate an icy fantasy character. Anatomical readability is substantially weaker: the ringed torso, crowded facial construction and fingerless appendages do not resolve the requested childlike crouch or pouncing gesture. Sparse surface detail and shallow shading further limit the result.","categories":[{"id":"fidelity","label":"Task & likeness","score":4},{"id":"structure","label":"Anatomy & construction","score":3},{"id":"space","label":"Space & placement","score":4},{"id":"light","label":"Light & reflections","score":4.5},{"id":"detail","label":"Purposeful detail","score":4},{"id":"artistry","label":"Artistry & character","score":4.5},{"id":"effects","label":"Materials & effects","score":4.5},{"id":"integrity","label":"Visual integrity","score":4}],"maximum":100},"taskComplexity":{"cognitive":2,"visual":4,"scale":10},"taskDescription":"Character scenes test expression, anatomy, materials and coherent visual construction.","promptSummary":"Character scenes test expression, anatomy, materials and coherent visual construction.","promptSummarized":true,"promptWithheld":true,"taskInstructionVariants":1}],"capabilities":{"version":"criteria-capabilities-v5","metrics":[{"id":"general","name":"General Intelligence","shortName":"General","description":"Construction, perspective, shadows, anatomy, purposeful detail and physical consistency."},{"id":"visual","name":"Visual Intelligence","shortName":"Visual","description":"Visual reconstruction, anatomy, materials, fine detail and artistic expression."}]},"gameResults":[{"modelId":"openai/gpt-6-luna","chainId":"foresight-self-improvement","family":"foresight","score":29.8,"cognitiveScore":29.8,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/a289497b7ededde5bf01.png","previewScore":{"t":7.158,"score":6,"turns":7,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/1c2c80f31ace24ba5d4c.svg","durationSeconds":12.146,"score":15,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/70174a3a4af5b0894444.svg","durationSeconds":8.964,"score":5,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/a231d6318d9288c03bb9.svg","durationSeconds":8.258,"score":6,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-bde96578ef3bcf8fccf1.json"},"rank":32},{"modelId":"openai/gpt-6-luna","chainId":"latch-five-levels","family":"latch","score":44,"cognitiveScore":44,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/0bbce04e7ea1a626ffb7.png","previewScore":{"t":5.659999974,"score":13,"turns":15,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/0031cf15899072407734.svg","durationSeconds":7.44,"score":8.8,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/066b4e4c7af0782323a1.svg","durationSeconds":11.34,"score":17,"outcome":"cleared","beaconsLeft":1},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/47ee1f20d73a57faaf3c.svg","durationSeconds":11.6,"score":15.8,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/6708d3512c6560199aa3.svg","durationSeconds":6.14,"score":13,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-f4b7a8bf42dc3db47787.json"},"rank":25},{"modelId":"openai/gpt-6.1-sol","chainId":"latch-five-levels","family":"latch","score":87.6,"cognitiveScore":87.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/8c86c49ee50186b50015.svg","durationSeconds":5.62,"score":10.2,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/6b20ec75d0ebb5180e21.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/1bc1e19b39450888020f.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/2805628348240a6011fc.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-4a8d1518abe5ed8723bc.json"},"rank":1},{"modelId":"openai/gpt-6.1-sol","chainId":"foresight-self-improvement","family":"foresight","score":62,"cognitiveScore":62,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/808eea277a9930c2cbc4.png","previewScore":{"t":13.008,"score":18,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/98560dab3c65072ac1ee.svg","durationSeconds":17.69,"score":26,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/b45049490f7e95a0e3d8.svg","durationSeconds":10.26,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/f34fd9f815f4c9282c15.svg","durationSeconds":14.508,"score":18,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-918c077dac42a5928046.json"},"rank":9},{"modelId":"cat-jam","chainId":"latch-five-levels","family":"latch","score":33.9,"cognitiveScore":33.9,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/f093966d5c051f3f47d0.png","previewScore":{"t":8.2599999866,"score":6,"turns":25,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/14b8b2bb595049f68f28.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/6226d2ac07c57c73ba0f.svg","durationSeconds":7.7,"score":10,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/d3d0710e61925df9627a.svg","durationSeconds":10.3,"score":10.8,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/e70865ba5221a2eace1e.svg","durationSeconds":8.74,"score":6,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-cdd09aed7e4a0ddd10f9.json"},"rank":33},{"modelId":"cat-jam","chainId":"foresight-self-improvement","family":"foresight","score":33.8,"cognitiveScore":33.8,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/207012e896af735a5a04.png","previewScore":{"t":12.508,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/feb46a7aa2278ec2bb81.svg","durationSeconds":8.964,"score":5,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/2e9f52add7b2c88026cc.svg","durationSeconds":10.26,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/ff24ae554774c156f0bb.svg","durationSeconds":13.608,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-7664046603359b60a35d.json"},"rank":26},{"modelId":"openai/gpt-6-astra","chainId":"foresight-self-improvement","family":"foresight","score":76.1,"cognitiveScore":76.1,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/a23e1713ce420337e6ef.png","previewScore":{"t":12.734,"score":18,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/bb9c8a135479e3d62622.svg","durationSeconds":17.416,"score":27,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/e5893b8b2c08526009d6.svg","durationSeconds":13.644,"score":11,"outcome":"cleared"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/2df6c5279ec52c72bbd4.svg","durationSeconds":14.234,"score":18,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-729b00b60eaae0918dfb.json"},"rank":2},{"modelId":"openai/gpt-6-astra","chainId":"latch-five-levels","family":"latch","score":87.6,"cognitiveScore":87.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/827a2df06d10d89423c9.svg","durationSeconds":5.62,"score":10.2,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/4115501ba403fc456d71.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/1bc1e19b39450888020f.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/16a51bb30db6dd1aa8c9.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-7773d04ee223aba6127a.json"},"rank":1},{"modelId":"anthropic/claude-sonnet-5.5","chainId":"foresight-self-improvement","family":"foresight","score":59.6,"cognitiveScore":59.6,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/5787d4aca32639aafe12.png","previewScore":{"t":12.734,"score":10,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/8b4328c3fb69d82dcbec.svg","durationSeconds":17.416,"score":25,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/b43ccec5ae15ec2c66a7.svg","durationSeconds":12.852,"score":10,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/4041e70f9e93f87b038e.svg","durationSeconds":14.234,"score":10,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-954ba298f82895bc3f8e.json"},"rank":11},{"modelId":"anthropic/claude-sonnet-5.5","chainId":"latch-five-levels","family":"latch","score":83.9,"cognitiveScore":83.9,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/0f981b65d08b0608496c.svg","durationSeconds":5.88,"score":10,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2ef1b80d5b5e63b85151.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/1bc1e19b39450888020f.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/2dfc40f48dbb2ca6fa8a.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-d4cb5078d80041e7db1f.json"},"rank":4},{"modelId":"google/gemma-4-26b-a4b-it","chainId":"foresight-self-improvement","family":"foresight","score":4.1,"cognitiveScore":4.1,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/c983fd594566d5f21f6d.png","previewScore":{"t":8.57,"score":2,"turns":9,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/b7a3927a1ca29cd1d662.svg","durationSeconds":8.374,"score":3,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/88774ffe69d303e951c9.svg","durationSeconds":8.1,"score":2,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/0487675c79411512fec2.svg","durationSeconds":9.67,"score":2,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-cbcd09f9816896130c58.json"},"rank":46},{"modelId":"google/gemma-4-26b-a4b-it","chainId":"latch-five-levels","family":"latch","score":4.5,"cognitiveScore":4.5,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/ce80ec01013fe61e8609.png","previewScore":{"t":11.900000040200002,"score":3,"turns":40,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/d166b1a3337865049cab.svg","durationSeconds":5.1,"score":3,"outcome":"cleared","beaconsLeft":1},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/17fb3f578f87055e8805.svg","durationSeconds":8.22,"score":6.2,"outcome":"ended-early"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/4e05250f075a9c2cda7f.svg","durationSeconds":11.34,"score":15.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/939d4cadd773feb610eb.svg","durationSeconds":2.5,"score":11.8,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/4ba7587ac166e7d85847.svg","durationSeconds":12.38,"score":3,"outcome":"out-of-turns"}],"detail":"/data/v1/benchmark-detail/replay-0bb5c1d05cf06d77b08c.json"},"rank":41},{"modelId":"tencent/hy4-preview","chainId":"foresight-self-improvement","family":"foresight","score":53.1,"cognitiveScore":53.1,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/16953d2a00159093c197.png","previewScore":{"t":11.046,"score":11,"turns":16,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/7016239d439726e18eb5.svg","durationSeconds":17.69,"score":26,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/d76921eb73d69776d382.svg","durationSeconds":9.056,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/5adaa4e44f7abefb22b1.svg","durationSeconds":12.146,"score":11,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-60a4f9eaf5cc511116b0.json"},"rank":14},{"modelId":"tencent/hy4-preview","chainId":"latch-five-levels","family":"latch","score":35.6,"cognitiveScore":35.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/eab84fa5d43738c5e3b6.png","previewScore":{"t":6.6999999834,"score":18.2,"turns":19,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/52ced9d86dfbee0da7f5.svg","durationSeconds":7.18,"score":6,"outcome":"ended-early"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/bc86a2d1041d83a39c7a.svg","durationSeconds":13.86,"score":18,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/a1ebef7d5da550337e86.svg","durationSeconds":11.6,"score":15.8,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/8d11c1724055aa36c72d.svg","durationSeconds":7.18,"score":18.2,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-ec2e109350a833020109.json"},"rank":31},{"modelId":"minimax/minimax-m3","chainId":"foresight-self-improvement","family":"foresight","score":22.1,"cognitiveScore":22.1,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/a289497b7ededde5bf01.png","previewScore":{"t":7.158,"score":6,"turns":7,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/1ac0a936512bb4776cb5.svg","durationSeconds":14.422,"score":17,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/4be59e8125078dc658c7.svg","durationSeconds":10.808,"score":3,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/a231d6318d9288c03bb9.svg","durationSeconds":8.258,"score":6,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-2f7463d1f1bdee9ea2d2.json"},"rank":38},{"modelId":"minimax/minimax-m3","chainId":"latch-five-levels","family":"latch","score":31.4,"cognitiveScore":31.4,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/a49f4f423f79b1873e7c.png","previewScore":{"t":4.619999985,"score":8.8,"turns":11,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/982d7c4be1653f458827.svg","durationSeconds":6.66,"score":9.4,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/ca9036fa1858a157a9ea.svg","durationSeconds":11.34,"score":0.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/a21a21f2a9868dc050df.svg","durationSeconds":7.18,"score":14.2,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/ff4b6ecbd72ae9355ed4.svg","durationSeconds":5.1,"score":8.8,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-672450784e9ca07f3a1d.json"},"rank":34},{"modelId":"inclusionai/ling-3.1-flash","chainId":"foresight-self-improvement","family":"foresight","score":43.1,"cognitiveScore":43.1,"coverage":{"completed":1,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/55ccbce16982139782b1.png","previewScore":{"t":10.182,"score":6,"turns":14,"limit":18},"previewLabel":"Last · Level 2/3","date":"2026-10-21","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":false,"pending":true},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/adbe019c66662b172a42.svg","durationSeconds":11.282,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":false}],"detail":"/data/v1/benchmark-detail/replay-5ad442a38a950c31aa07.json"},"rank":42},{"modelId":"inclusionai/ling-3.1-flash","chainId":"latch-five-levels","family":"latch","score":45.9,"cognitiveScore":45.9,"coverage":{"completed":3,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/15ee31f907f0327b39cd.png","previewScore":{"t":7.220000018,"score":9.8,"turns":21,"limit":40},"previewLabel":"Last · Level 4/5","date":"2026-10-21","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/96ac30f5dc846da4883a.svg","durationSeconds":10.22,"score":9.4,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":false,"pending":true},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/68c3dadc156eb49bbc2e.svg","durationSeconds":7.7,"score":9.8,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":false}],"detail":"/data/v1/benchmark-detail/replay-4fa281b749184383aef8.json"},"rank":36},{"modelId":"openai/gpt-5.6-luna","chainId":"foresight-self-improvement","family":"foresight","score":25.1,"cognitiveScore":25.1,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/4fc5cdbca5ca42f469f5.png","previewScore":{"t":10.614,"score":8,"turns":15,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/7e9e6433bd24214a4844.svg","durationSeconds":8.69,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/bef204297827265e603f.svg","durationSeconds":9.396,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/79f8cfcf6f638c448ed9.svg","durationSeconds":11.714,"score":8,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-6f9662a24608e259f77d.json"},"rank":35},{"modelId":"openai/gpt-5.6-luna","chainId":"latch-five-levels","family":"latch","score":52.9,"cognitiveScore":52.9,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/ec2ab747fc0a6a01ee2e.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/a74f7ea847209b46b27e.svg","durationSeconds":7.7,"score":8.6,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/519e5c7016a8e1e6740f.svg","durationSeconds":13.86,"score":18,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/91f7f61fced710cafccd.svg","durationSeconds":10.04,"score":17,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/8014b6e5ea4907d5c009.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-f6cf76b8bdc0fa72273e.json"},"rank":21},{"modelId":"meta/muse-spark-open","chainId":"foresight-self-improvement","family":"foresight","score":29.4,"cognitiveScore":29.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/4fc5cdbca5ca42f469f5.png","previewScore":{"t":10.614,"score":8,"turns":15,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/356d421180bae8b2776d.svg","durationSeconds":8.69,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/3dae4fe5c0ac26077758.svg","durationSeconds":11.124,"score":9,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/79f8cfcf6f638c448ed9.svg","durationSeconds":11.714,"score":8,"outcome":"out-of-lives","invalid":1}],"detail":"/data/v1/benchmark-detail/replay-7956698d6af76172c458.json"},"rank":34},{"modelId":"meta/muse-spark-open","chainId":"latch-five-levels","family":"latch","score":75.7,"cognitiveScore":75.7,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/2e68d0d333415f7d1b6f.png","previewScore":{"t":9.820000004799999,"score":19.6,"turns":32,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/f894c31cd335afa76eae.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/206d34fd1f80d6d92e44.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/245039ec25614c967d4d.svg","durationSeconds":11.6,"score":17.8,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/12e174c5be75f513a386.svg","durationSeconds":12.82,"score":19.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-d4e45ec75d6798a4362b.json"},"rank":12},{"modelId":"openai/gpt-5.6-sol","chainId":"foresight-self-improvement","family":"foresight","score":62.9,"cognitiveScore":62.9,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/2919aeea83ec4c804b26.png","previewScore":{"t":12.026,"score":11,"turns":17,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/f1682e5fb12e307f65bc.svg","durationSeconds":17.416,"score":27,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/5d389bd1c7d07e3ef5da.svg","durationSeconds":12.852,"score":10,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/f98ed64ef79a01619be2.svg","durationSeconds":13.126,"score":11,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-e59883d71c29c249c6ea.json"},"rank":8},{"modelId":"openai/gpt-5.6-sol","chainId":"latch-five-levels","family":"latch","score":83.6,"cognitiveScore":83.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/ec2ab747fc0a6a01ee2e.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/0f0f6e54de0b7f8edbc7.svg","durationSeconds":5.62,"score":10.2,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/4359b0183cfb63932954.svg","durationSeconds":12.82,"score":18.8,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/dd41c8c49c5a91b58bd8.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/df03bd090d7b1fed1112.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-9f757a5b232b2b9a1221.json"},"rank":5},{"modelId":"openai/gpt-5.6-terra","chainId":"foresight-self-improvement","family":"foresight","score":30.1,"cognitiveScore":30.1,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/a289497b7ededde5bf01.png","previewScore":{"t":7.158,"score":6,"turns":7,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-15","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/caf73e4f6c27e386bbec.svg","durationSeconds":8.964,"score":5,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/4ceb6e1660946d16bd26.svg","durationSeconds":13.284,"score":9,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/a231d6318d9288c03bb9.svg","durationSeconds":8.258,"score":6,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-515f922b7b00d95c394e.json"},"rank":31},{"modelId":"openai/gpt-5.6-terra","chainId":"latch-five-levels","family":"latch","score":80,"cognitiveScore":80,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/2e68d0d333415f7d1b6f.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/e379b17375af0adbd7c1.svg","durationSeconds":9.96,"score":9.6,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/a07b51f899938f5e961d.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/1bc1e19b39450888020f.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/0925c10dbbb08d5ae45b.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-6b36824ef6cfe01df069.json"},"rank":9},{"modelId":"space-bunny-alpha","chainId":"foresight-self-improvement","family":"foresight","score":35.8,"cognitiveScore":35.8,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/40770a77f26313f4d3ed.png","previewScore":{"t":12.342,"score":8,"turns":19,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/fe3480084c9b3637f67d.svg","durationSeconds":13.284,"score":16,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/bf7b7596ed859689d0f9.svg","durationSeconds":9.396,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/d1318ba7eb0bf08984ce.svg","durationSeconds":13.442,"score":8,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-0421ece81b998b95dea0.json"},"rank":24},{"modelId":"space-bunny-alpha","chainId":"latch-five-levels","family":"latch","score":34.4,"cognitiveScore":34.4,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/5727ec1668a60efd37f2.png","previewScore":{"t":6.1800000138,"score":12.6,"turns":17,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/3c42d78cdd10134bc897.svg","durationSeconds":6.32,"score":5,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/d030183fd2183d724eab.svg","durationSeconds":10.74,"score":9,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2bb1870abe3ec025d1b4.svg","durationSeconds":10.04,"score":16.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/cc1876eeba0574430d0d.svg","durationSeconds":10.82,"score":10.4,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/3f91f0f61dc79db1dd98.svg","durationSeconds":6.66,"score":12.6,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-3ab04ee123b3c38a9eea.json"},"rank":32},{"modelId":"pixel-canary","chainId":"latch-five-levels","family":"latch","score":100,"cognitiveScore":100,"coverage":{"completed":1,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/8de9ad39a066d50221b3.png","previewScore":{"t":3.0600000299999994,"score":5.2,"turns":6,"limit":12},"previewLabel":"Last · Level 1/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":false,"pending":true},{"label":"Level 3 of 5 · Relay vault","available":false,"pending":true},{"label":"Level 4 of 5 · Passing bay","available":false},{"label":"Level 5 of 5 · Second thoughts","available":false}],"detail":"/data/v1/benchmark-detail/replay-0138d08f6b0d799f5318.json"},"rank":40},{"modelId":"anthropic/claude-haiku-5.5","chainId":"foresight-self-improvement","family":"foresight","score":34.4,"cognitiveScore":34.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/807eec82d4bb4ca3436c.png","previewScore":{"t":15.248,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/aad987507bb2bccb8521.svg","durationSeconds":8.532,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/59d97c2ea8eef7e22ea1.svg","durationSeconds":11.398,"score":7,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/9a9237cb44e09f5e8576.svg","durationSeconds":16.348,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-f5605337ba9c825d664d.json"},"rank":25},{"modelId":"anthropic/claude-haiku-5.5","chainId":"latch-five-levels","family":"latch","score":78.3,"cognitiveScore":78.3,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/bd048541c397e0fef30f.svg","durationSeconds":6.32,"score":5,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/f2a9ee88c718c47dc093.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/8fd3e14df8870aa0b729.svg","durationSeconds":12.82,"score":18.8,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/dd41c8c49c5a91b58bd8.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/2805628348240a6011fc.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-99a5b33218d24deb38d3.json"},"rank":10},{"modelId":"anthropic/claude-opus-5.5","chainId":"foresight-self-improvement","family":"foresight","score":66.3,"cognitiveScore":66.3,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/f8acf57f0044c86328dd.png","previewScore":{"t":13.33,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/bb9c8a135479e3d62622.svg","durationSeconds":17.416,"score":27,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/fba67a378cd038987d49.svg","durationSeconds":13.018,"score":11,"outcome":"cleared"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/0c4c51f4709611304f7a.svg","durationSeconds":14.43,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-fee3b8cc6bef154bb8d7.json"},"rank":5},{"modelId":"anthropic/claude-opus-5.5","chainId":"latch-five-levels","family":"latch","score":87.6,"cognitiveScore":87.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/8c86c49ee50186b50015.svg","durationSeconds":5.62,"score":10.2,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/6366742523a93db8954c.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/dd41c8c49c5a91b58bd8.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/2dfc40f48dbb2ca6fa8a.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-0d505ac342636c167f16.json"},"rank":1},{"modelId":"glyph-cluster","chainId":"foresight-self-improvement","family":"foresight","score":6.4,"cognitiveScore":6.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/8ebb51dcfcdab513f822.png","previewScore":{"t":6.568,"score":3,"turns":5,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/b7a3927a1ca29cd1d662.svg","durationSeconds":8.374,"score":3,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/541dfd93459bd75e8e51.svg","durationSeconds":9.512,"score":3,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/fd1a0fa1303968009837.svg","durationSeconds":7.668,"score":3,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-ef01482e228d8e2e0276.json"},"rank":44},{"modelId":"glyph-cluster","chainId":"latch-five-levels","family":"latch","score":3.6,"cognitiveScore":3.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/724b5f8f342cefb1f76c.png","previewScore":{"t":8.2599999866,"score":9,"turns":25,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/91b1d69b4092c19e5a52.svg","durationSeconds":8.22,"score":2,"outcome":"out-of-turns"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/eaad8fd05df1ee07965f.svg","durationSeconds":11.34,"score":4.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/cda374eae83d8f866859.svg","durationSeconds":12.38,"score":1,"outcome":"out-of-turns"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/4534814701a6cbf79d90.svg","durationSeconds":8.74,"score":9,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-2c5459d13a4c2cf49b2a.json"},"rank":44},{"modelId":"qwen/qwen3.8-flash","chainId":"foresight-self-improvement","family":"foresight","score":29.7,"cognitiveScore":29.7,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/4c103ffa1a6184a560e6.png","previewScore":{"t":9.002,"score":5,"turns":10,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-23","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/98560dab3c65072ac1ee.svg","durationSeconds":17.69,"score":26,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/32f27e810c17396fe389.svg","durationSeconds":7.328,"score":3,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/23e48e0923df9ad6bbc0.svg","durationSeconds":10.102,"score":5,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-e5de1e89a0a52574bf0d.json"},"rank":33},{"modelId":"qwen/qwen3.6-27b","chainId":"foresight-self-improvement","family":"foresight","score":37.8,"cognitiveScore":37.8,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/1919860d30fa4f4f1388.png","previewScore":{"t":13.764,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/aecf075fb0de34d02b0b.svg","durationSeconds":9.92,"score":9,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/5058a6c099d0d388c67d.svg","durationSeconds":9.396,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/584020d7eee48a786443.svg","durationSeconds":14.864,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-176f4d833b4850fa3f50.json"},"rank":23},{"modelId":"qwen/qwen3.8-flash","chainId":"latch-five-levels","family":"latch","score":69.8,"cognitiveScore":69.8,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/2e68d0d333415f7d1b6f.png","previewScore":{"t":9.299999988,"score":20,"turns":30,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-19","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/f5f59f7fbc21729abc00.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/b8849d72eef89848872b.svg","durationSeconds":12.82,"score":18.8,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/962f0ed2823d9fd911b5.svg","durationSeconds":9.78,"score":17.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/6493f09abf90acd293d4.svg","durationSeconds":12.3,"score":20,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-6524297f46d692e7bd2c.json"},"rank":16},{"modelId":"qwen/qwen3.6-27b","chainId":"latch-five-levels","family":"latch","score":28.5,"cognitiveScore":28.5,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/0cd1770ece1253acd115.png","previewScore":{"t":8.2599999866,"score":11,"turns":25,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/49a2b2585968d7a94372.svg","durationSeconds":6.58,"score":4.8,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/f41d0377dae2a2f376d1.svg","durationSeconds":7.44,"score":8.8,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/7ba00f22e3b035d2dfcb.svg","durationSeconds":9.78,"score":16.4,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/2abbf894ef06f7cff99b.svg","durationSeconds":8.22,"score":13.4,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/6ea3e628d4a5a7249d7e.svg","durationSeconds":8.74,"score":11,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-3ce1f5c048a28105a91c.json"},"rank":35},{"modelId":"xiaomi/mimo-v2.6-flash","chainId":"foresight-self-improvement","family":"foresight","score":71.3,"cognitiveScore":71.3,"coverage":{"completed":1,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/4459f9e2d6ec62880fdd.png","previewScore":{"t":10.456,"score":12,"turns":14,"limit":18},"previewLabel":"Last · Level 2/3","date":"2026-10-19","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":false,"pending":true},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/3633163a02b3c7ecb3fb.svg","durationSeconds":11.556,"score":12,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":false}],"detail":"/data/v1/benchmark-detail/replay-1df1fe0511cd726f9048.json"},"rank":36},{"modelId":"xiaomi/mimo-v2.6-flash","chainId":"latch-five-levels","family":"latch","score":66.3,"cognitiveScore":66.3,"coverage":{"completed":3,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/4eb4c7200e6fab9d3340.png","previewScore":{"t":10.859999950799999,"score":18,"turns":36,"limit":36},"previewLabel":"Last · Level 3/5","date":"2026-10-19","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/a0078dac4fdf471df4c4.svg","durationSeconds":9.96,"score":9.6,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2dd9d2f5f59e3ec9b068.svg","durationSeconds":13.86,"score":18,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":false,"pending":true},{"label":"Level 5 of 5 · Second thoughts","available":false}],"detail":"/data/v1/benchmark-detail/replay-d90b7c88fa4f54331d37.json"},"rank":27},{"modelId":"openai/gpt-6-sol","chainId":"foresight-self-improvement","family":"foresight","score":30.4,"cognitiveScore":30.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/a289497b7ededde5bf01.png","previewScore":{"t":7.158,"score":6,"turns":7,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/d30afcff88688e988aa7.svg","durationSeconds":14.422,"score":17,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/3643485c01563b51b97b.svg","durationSeconds":10.26,"score":5,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/a231d6318d9288c03bb9.svg","durationSeconds":8.258,"score":6,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-903c7cdc65d6dd52c650.json"},"rank":29},{"modelId":"ox-alpha","chainId":"foresight-self-improvement","family":"foresight","score":49,"cognitiveScore":49,"coverage":{"completed":2,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/121d4b86866d7e96630c.png","previewScore":{"t":12.12,"score":16,"turns":20,"limit":20},"previewLabel":"Last · Level 3/3","date":"2026-10-23","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/aecf075fb0de34d02b0b.svg","durationSeconds":9.92,"score":9,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":false,"pending":true},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/eeac6afd9bf74d7acf9d.svg","durationSeconds":13.528,"score":16,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-4d742d2a1a1e877eb1e3.json"},"rank":27},{"modelId":"ox-alpha","chainId":"latch-five-levels","family":"latch","score":4,"cognitiveScore":4,"coverage":{"completed":3,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/071934e91f31bf71888b.png","previewScore":{"t":10.3400000222,"score":0.6,"turns":33,"limit":36},"previewLabel":"Last · Level 3/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/073cd8104a7cc2196c38.svg","durationSeconds":7.1,"score":4.4,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/542d6e08e5922fd0cbeb.svg","durationSeconds":6.92,"score":5.2,"outcome":"ended-early"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/f2031712afa905f0de49.svg","durationSeconds":10.82,"score":0.6,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":false,"pending":true},{"label":"Level 5 of 5 · Second thoughts","available":false}],"detail":"/data/v1/benchmark-detail/replay-26bec39e538bc555daf1.json"},"rank":48},{"modelId":"qwen/qwen3.5-9b","chainId":"foresight-self-improvement","family":"foresight","score":12.8,"cognitiveScore":12.8,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/e97e17369b899dd42d5a.png","previewScore":{"t":8.686,"score":2,"turns":8,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/f6376d6be5e4b7510a18.svg","durationSeconds":9.512,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/d4a23b57ef0f20cdbfb1.svg","durationSeconds":10.376,"score":4,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/050c66b49d403ca8a9d0.svg","durationSeconds":9.786,"score":2,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-4c1ed410fb268a4e50da.json"},"rank":43},{"modelId":"openai/gpt-6-sol","chainId":"latch-five-levels","family":"latch","score":76.6,"cognitiveScore":76.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/275b571dca77e7e93ec1.svg","durationSeconds":9.96,"score":9.6,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/86cc4585d5737225f457.svg","durationSeconds":13.34,"score":18.4,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/1bc1e19b39450888020f.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/2dfc40f48dbb2ca6fa8a.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-cfa6e769840851c6f713.json"},"rank":11},{"modelId":"qwen/qwen3.5-9b","chainId":"latch-five-levels","family":"latch","score":3.8,"cognitiveScore":3.8,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/3602d2d36d551867fc18.png","previewScore":{"t":10.3400000222,"score":4.4,"turns":33,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/f429a850e95cb2f4f656.svg","durationSeconds":6.32,"score":5,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/01a923b1275bcaa2e21b.svg","durationSeconds":6.14,"score":3.8,"outcome":"ended-early"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2cf2559faa7fdf732e73.svg","durationSeconds":10.04,"score":1.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/09ab24747f73b5242836.svg","durationSeconds":12.38,"score":0.2,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/0edea3c284dd99b8a06a.svg","durationSeconds":10.82,"score":4.4,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-8144fc90b3e9beb352c4.json"},"rank":43},{"modelId":"anthropic/claude-fable-5.1","chainId":"foresight-self-improvement","family":"foresight","score":34.2,"cognitiveScore":34.2,"coverage":{"completed":2,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/4e7792370abf5dd3527d.png","previewScore":{"t":11.478,"score":8,"turns":17,"limit":20},"previewLabel":"Last · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/87c8fa900e1ac3828af7.svg","durationSeconds":15.104,"score":21,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":false,"pending":true},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/cfe75e813dff6199ef19.svg","durationSeconds":12.578,"score":8,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-0e1f96c058d7d2a00437.json"},"rank":37},{"modelId":"anthropic/claude-fable-5.1","chainId":"latch-five-levels","family":"latch","score":80.2,"cognitiveScore":80.2,"coverage":{"completed":3,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"Last · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":false,"pending":true},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/b18b6ce461a3bd67adcc.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":false,"pending":true},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/dd41c8c49c5a91b58bd8.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/16a51bb30db6dd1aa8c9.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-575313e1716a04d2841f.json"},"rank":24},{"modelId":"deepseek/deepseek-v4-pro-0813","chainId":"foresight-self-improvement","family":"foresight","score":42.7,"cognitiveScore":42.7,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/947e528420b742fd4e61.png","previewScore":{"t":13.49,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/97cd429774e143fab7d0.svg","durationSeconds":13.126,"score":14,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/23ab77cdbe2f6c3cc782.svg","durationSeconds":9.396,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/100dffaf2fab02e9c400.svg","durationSeconds":14.59,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-26401b4b33b785ab7896.json"},"rank":21},{"modelId":"deepseek/deepseek-v4-pro-0813","chainId":"latch-five-levels","family":"latch","score":71.7,"cognitiveScore":71.7,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/c048bc501d5699a06005.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/20d235ac814fa5159119.svg","durationSeconds":12.82,"score":18.8,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/309af1fc3076ddab84b5.svg","durationSeconds":12.38,"score":17.2,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/d8d86aa930f884095ca2.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-72b1f37ee5c8dc4698a7.json"},"rank":15},{"modelId":"deepseek/deepseek-v4.1-flash","chainId":"foresight-self-improvement","family":"foresight","score":55.4,"cognitiveScore":55.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/d406959b553cc83d9109.png","previewScore":{"t":13.008,"score":18,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/1e95b346c5e3ea34391a.svg","durationSeconds":13.01,"score":16,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/f20e5f97074a73200410.svg","durationSeconds":9.92,"score":7,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/2f9b795cf83635cd33c3.svg","durationSeconds":14.508,"score":18,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-33ac1e28b42f285a5e3e.json"},"rank":12},{"modelId":"deepseek/deepseek-v4.1-flash","chainId":"latch-five-levels","family":"latch","score":65.2,"cognitiveScore":65.2,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/3f36d73cd14f713b816d.svg","durationSeconds":10.74,"score":9,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/c05414d77b14512c0aaf.svg","durationSeconds":8.74,"score":17.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/1bc1e19b39450888020f.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/2805628348240a6011fc.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-1fc51a93a164cc204afa.json"},"rank":19},{"modelId":"google/gemma-4-31b-it","chainId":"foresight-self-improvement","family":"foresight","score":30.2,"cognitiveScore":30.2,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/d83fd5ae39ad79e7fb60.png","previewScore":{"t":12.258,"score":8,"turns":15,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/aecf075fb0de34d02b0b.svg","durationSeconds":9.92,"score":9,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/d76921eb73d69776d382.svg","durationSeconds":9.056,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/9cac3d1280bcc72ef331.svg","durationSeconds":13.358,"score":8,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-761f82d6fbffbba994a1.json"},"rank":30},{"modelId":"google/gemma-4-31b-it","chainId":"latch-five-levels","family":"latch","score":26.6,"cognitiveScore":26.6,"coverage":{"completed":4,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29e338241238487aa357.png","previewScore":{"t":11.900000040200002,"score":2.2,"turns":39,"limit":40},"previewLabel":"Last · Level 4/5","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/ec2ab747fc0a6a01ee2e.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/2d0328fb582b90ac317c.svg","durationSeconds":10.74,"score":9,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/e1ca30e842fc43092789.svg","durationSeconds":11.34,"score":13.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/2b4749a2abd60f418511.svg","durationSeconds":12.38,"score":2.2,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":false}],"detail":"/data/v1/benchmark-detail/replay-11e6e621cd8a4c58845e.json"},"rank":39},{"modelId":"z-ai/glm-5.3","chainId":"foresight-self-improvement","family":"foresight","score":59.1,"cognitiveScore":59.1,"coverage":{"completed":2,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/33cafede7d552736c13d.png","previewScore":{"t":9.276,"score":6,"turns":10,"limit":18},"previewLabel":"Last · Level 2/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/f1682e5fb12e307f65bc.svg","durationSeconds":17.416,"score":27,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/9aa7edbd09c9bc2e3d58.svg","durationSeconds":10.376,"score":6,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":false,"pending":true}],"detail":"/data/v1/benchmark-detail/replay-724472c549ff86de0e7c.json"},"rank":22},{"modelId":"z-ai/glm-5.3","chainId":"latch-five-levels","family":"latch","score":25.6,"cognitiveScore":25.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6df1a259245ba21c5c93.png","previewScore":{"t":5.400000018,"score":12.2,"turns":14,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/38483439cc32e21b90e7.svg","durationSeconds":10.74,"score":9,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/b4eebe873c48b0c5d24c.svg","durationSeconds":10.82,"score":7.6,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/dd31b1c370b825695ec9.svg","durationSeconds":11.08,"score":3.2,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/5373a90fa5b20487257c.svg","durationSeconds":5.88,"score":12.2,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-b23eb611d83b744a8736.json"},"rank":37},{"modelId":"google/gemini-3.8-flash","chainId":"foresight-self-improvement","family":"foresight","score":45.4,"cognitiveScore":45.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/2919aeea83ec4c804b26.png","previewScore":{"t":12.026,"score":11,"turns":17,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/bd90780471052ea2645e.svg","durationSeconds":8.964,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/a3411e1410220e9ceea8.svg","durationSeconds":13.096,"score":10,"outcome":"cleared"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/f98ed64ef79a01619be2.svg","durationSeconds":13.126,"score":11,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-8738719b34135d1dd2be.json"},"rank":20},{"modelId":"google/gemini-3.8-flash","chainId":"latch-five-levels","family":"latch","score":72.9,"cognitiveScore":72.9,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/8542a0c40f1211f1817a.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/dfe5cde8a6c7034e4883.svg","durationSeconds":12.82,"score":18.8,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/b10226d5d142096a3c02.svg","durationSeconds":11.6,"score":17.8,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/c7cfd9efe4e476deeaeb.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-6dbb3979641c1d70ff97.json"},"rank":14},{"modelId":"microsoft/mai-thinking-1","chainId":"foresight-self-improvement","family":"foresight","score":6.4,"cognitiveScore":6.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/ee0372776f93af5adf73.png","previewScore":{"t":9.234,"score":1,"turns":8,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/b7a3927a1ca29cd1d662.svg","durationSeconds":8.374,"score":3,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/385d70e71b60d4b46fe1.svg","durationSeconds":9.08,"score":3,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/95ba9ea2a3723c794e69.svg","durationSeconds":10.334,"score":1,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-b481669d8fdd77109da0.json"},"rank":44},{"modelId":"microsoft/mai-thinking-1","chainId":"latch-five-levels","family":"latch","score":3.6,"cognitiveScore":3.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/cc6384cb313a95a15b9d.png","previewScore":{"t":5.920000000000001,"score":11.8,"turns":16,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/fcfad28219d6ac8b2706.svg","durationSeconds":8.22,"score":2,"outcome":"out-of-turns"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/b91d871dc5c0ee219ac9.svg","durationSeconds":11.34,"score":2.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/a64730feada4294621c4.svg","durationSeconds":8.74,"score":3,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/8872cfc70116bee5cd3c.svg","durationSeconds":6.4,"score":11.8,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-d0ced2396cf66f8e4ad4.json"},"rank":44},{"modelId":"mistralai/mistral-large-4-0","chainId":"foresight-self-improvement","family":"foresight","score":22,"cognitiveScore":22,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/b936ef141ec943120b54.png","previewScore":{"t":8.138,"score":4,"turns":8,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-19","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/8b8c64369d6d35322fce.svg","durationSeconds":6.962,"score":1,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/de1f091bd08dbf61755d.svg","durationSeconds":10.418,"score":9,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/af98ee5b8ad5f00c89ae.svg","durationSeconds":9.238,"score":4,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-3bf64969bcd3b5e0de71.json"},"rank":39},{"modelId":"mistralai/mistral-large-4-0","chainId":"latch-five-levels","family":"latch","score":49.4,"cognitiveScore":49.4,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":9.56000004,"score":19.8,"turns":31,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-22","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/ae9de96a5f226d69b67d.svg","durationSeconds":9.96,"score":9.6,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/ac618a51cc9c11d69b24.svg","durationSeconds":10.56,"score":15.8,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/513764a409bfaa3753dd.svg","durationSeconds":11.08,"score":16.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/b4ec304c095740bddbfa.svg","durationSeconds":12.56,"score":19.8,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-591e4ca8090e288693fe.json"},"rank":23},{"modelId":"meta/muse-spark-1.3","chainId":"foresight-self-improvement","family":"foresight","score":64.3,"cognitiveScore":64.3,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/2919aeea83ec4c804b26.png","previewScore":{"t":12.026,"score":11,"turns":17,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/c51232857c13b4631492.svg","durationSeconds":17.416,"score":26,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/a3411e1410220e9ceea8.svg","durationSeconds":13.096,"score":10,"outcome":"cleared"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/f98ed64ef79a01619be2.svg","durationSeconds":13.126,"score":11,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-118609a9fbc768160187.json"},"rank":6},{"modelId":"meta/muse-spark-1.3","chainId":"latch-five-levels","family":"latch","score":72.4,"cognitiveScore":72.4,"coverage":{"completed":3,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/4f6c0350e1d704c5c509.png","previewScore":{"t":9.820000004799999,"score":18.8,"turns":32,"limit":36},"previewLabel":"Last · Level 3/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/1df5fea57c0d382b52b4.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/37ed5145b27e7de17515.svg","durationSeconds":12.82,"score":18.8,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":false,"pending":true},{"label":"Level 5 of 5 · Second thoughts","available":false}],"detail":"/data/v1/benchmark-detail/replay-857f73cab2cf9a0c8d1d.json"},"rank":26},{"modelId":"nvidia/nemotron-3-ultra-550b-a55b","chainId":"foresight-self-improvement","family":"foresight","score":16.9,"cognitiveScore":16.9,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/31227e08eaec12a80fa3.png","previewScore":{"t":12.3,"score":8,"turns":17,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/cb72b8526d0f37c28e8b.svg","durationSeconds":8.532,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/db9ac36c2451d99c56f4.svg","durationSeconds":9.604,"score":4,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/e6a307b0df713097ff27.svg","durationSeconds":13.4,"score":8,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-b43b8675075e24d95d63.json"},"rank":41},{"modelId":"nvidia/nemotron-3-ultra-550b-a55b","chainId":"latch-five-levels","family":"latch","score":5,"cognitiveScore":5,"coverage":{"completed":4,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/a9b3cf3fa341f98eade1.png","previewScore":{"t":7.220000018,"score":13.8,"turns":21,"limit":40},"previewLabel":"Last · Level 4/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/cbefa5ba8e872d6cb411.svg","durationSeconds":6.92,"score":3.2,"outcome":"ended-early"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/bdf8eb1919a30c708ab0.svg","durationSeconds":9.78,"score":7.4,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/d96f52af8464dc703c89.svg","durationSeconds":7.7,"score":13.8,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":false}],"detail":"/data/v1/benchmark-detail/replay-a36b333f8cc014b8bad3.json"},"rank":42},{"modelId":"qwen/qwen3.8-27B","chainId":"foresight-self-improvement","family":"foresight","score":63,"cognitiveScore":63,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/16953d2a00159093c197.png","previewScore":{"t":11.046,"score":11,"turns":16,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/7016239d439726e18eb5.svg","durationSeconds":17.69,"score":26,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/0318ab3fb5f6fdedfd8f.svg","durationSeconds":11.556,"score":12,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/5adaa4e44f7abefb22b1.svg","durationSeconds":12.146,"score":11,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-7bfc7554b40718963a68.json"},"rank":7},{"modelId":"qwen/qwen3.8-27B","chainId":"latch-five-levels","family":"latch","score":38.6,"cognitiveScore":38.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/042c96f9dd8e27ce09bc.png","previewScore":{"t":9.820000018000002,"score":15.8,"turns":31,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/b3d166a5dcaf23f291e7.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/19e3b7f865441e164bb5.svg","durationSeconds":10.04,"score":11.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/625cd6c48e2331a963f6.svg","durationSeconds":12.38,"score":13.2,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/9ccc34dd2cd32a45d6c1.svg","durationSeconds":10.3,"score":15.8,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-2ad92f85128faaaab89e.json"},"rank":28},{"modelId":"qwen/qwen3.6-35b-a3b","chainId":"foresight-self-improvement","family":"foresight","score":20.7,"cognitiveScore":20.7,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/93e57465c038d139c030.png","previewScore":{"t":14.152,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/cb72b8526d0f37c28e8b.svg","durationSeconds":8.532,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/6d6da85aee2e8f0740fa.svg","durationSeconds":9.512,"score":3,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/b7d10f80948b8640f26b.svg","durationSeconds":15.252,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-e0260aaeb275002ff148.json"},"rank":40},{"modelId":"qwen/qwen3.6-35b-a3b","chainId":"latch-five-levels","family":"latch","score":3.6,"cognitiveScore":3.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/30eef7780a5dd487d410.png","previewScore":{"t":11.120000004,"score":7.8,"turns":36,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/ec2ab747fc0a6a01ee2e.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/8e248931c6c5b12d01b4.svg","durationSeconds":8.22,"score":1,"outcome":"out-of-turns"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/b9510766865b45abc674.svg","durationSeconds":11.34,"score":0.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/48d246fee0138532e10f.svg","durationSeconds":12.38,"score":3,"outcome":"out-of-turns"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/956a028d380bdb40ba02.svg","durationSeconds":11.6,"score":7.8,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-26566417e0167cc2e096.json"},"rank":44},{"modelId":"qwen/qwen3.8-max-0902","chainId":"foresight-self-improvement","family":"foresight","score":72.4,"cognitiveScore":72.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/fb581eeeed224268a050.png","previewScore":{"t":12.668,"score":16,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/bb9c8a135479e3d62622.svg","durationSeconds":17.416,"score":27,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/e073d510b290e409ccc7.svg","durationSeconds":11.556,"score":12,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/e2d8acc53f4c0361aaad.svg","durationSeconds":13.768,"score":16,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-57560d3a25439a532d2d.json"},"rank":4},{"modelId":"qwen/qwen3.8-max-0902","chainId":"latch-five-levels","family":"latch","score":82.6,"cognitiveScore":82.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/c9736767e7a5e201b447.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/f96bb76d96ee3ce50221.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/525ec8dffa992e7118b9.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/72ebb128fc07c5d617d5.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-69105ef2abb859e6f152.json"},"rank":6},{"modelId":"x-ai/grok-4.7","chainId":"foresight-self-improvement","family":"foresight","score":72.6,"cognitiveScore":72.6,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/0bfa5e467d3bc6a576fb.png","previewScore":{"t":12.734,"score":16,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/87f023e33dcba56cf181.svg","durationSeconds":17.69,"score":25,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/fe58ef9f97f1d2eeaca9.svg","durationSeconds":13.096,"score":13,"outcome":"cleared"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/886efe8104cc43911675.svg","durationSeconds":14.234,"score":16,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-1a72eb23d4a2c874c852.json"},"rank":3},{"modelId":"x-ai/grok-4.7","chainId":"latch-five-levels","family":"latch","score":65.4,"cognitiveScore":65.4,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":8.7800000046,"score":20.4,"turns":28,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/1df5fea57c0d382b52b4.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/f8fb1703d5d32ec6c885.svg","durationSeconds":9.52,"score":11.6,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/a96cc83ea445eb81a64a.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/3ff924e6cdfd287bf496.svg","durationSeconds":11.78,"score":20.4,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-a16ec23bec3186ecbe3a.json"},"rank":18},{"modelId":"google/gemini-4-argon","chainId":"foresight-self-improvement","family":"foresight","score":48.6,"cognitiveScore":48.6,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/a289497b7ededde5bf01.png","previewScore":{"t":7.158,"score":6,"turns":7,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/98560dab3c65072ac1ee.svg","durationSeconds":17.69,"score":26,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/b7f9ec893a041f45de85.svg","durationSeconds":10.692,"score":9,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/a231d6318d9288c03bb9.svg","durationSeconds":8.258,"score":6,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-aed579e9ba876f03f054.json"},"rank":19},{"modelId":"google/gemini-4-argon","chainId":"latch-five-levels","family":"latch","score":82.6,"cognitiveScore":82.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/29aa36983c93b6703c2d.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/ec2ab747fc0a6a01ee2e.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/e782366e5c65dcb1e0dd.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2ef1b80d5b5e63b85151.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/dd41c8c49c5a91b58bd8.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/16a51bb30db6dd1aa8c9.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-9afba30bf0d22823879e.json"},"rank":6},{"modelId":"moonshotai/kimi-k3","chainId":"foresight-self-improvement","family":"foresight","score":52.7,"cognitiveScore":52.7,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/2919aeea83ec4c804b26.png","previewScore":{"t":12.026,"score":11,"turns":17,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/bef3ae84185a58545fc0.svg","durationSeconds":16.466,"score":17,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/6bf3ee023dc7fdc2d86f.svg","durationSeconds":12.852,"score":10,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/f98ed64ef79a01619be2.svg","durationSeconds":13.126,"score":11,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-1839eaae4ee2e8faf220.json"},"rank":16},{"modelId":"moonshotai/kimi-k3","chainId":"latch-five-levels","family":"latch","score":57.7,"cognitiveScore":57.7,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/3b87ee5c6ae41e28ea2f.png","previewScore":{"t":2.8000000112,"score":14.2,"turns":4,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/d819e24f756e3223595c.svg","durationSeconds":9.96,"score":9.6,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/86ba3f0d639189c57ab8.svg","durationSeconds":7.44,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/25d9b38f97181d95de1f.svg","durationSeconds":11.08,"score":14.2,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/e102ea820a1e0ecdb345.svg","durationSeconds":3.28,"score":14.2,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-09719792f4e2e0c87a32.json"},"rank":20},{"modelId":"meta/muse-wonder","chainId":"foresight-self-improvement","family":"foresight","score":52.3,"cognitiveScore":52.3,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/2919aeea83ec4c804b26.png","previewScore":{"t":12.026,"score":11,"turns":17,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/fe3480084c9b3637f67d.svg","durationSeconds":13.284,"score":16,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/280a6435489f042592e0.svg","durationSeconds":12.852,"score":10,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/f98ed64ef79a01619be2.svg","durationSeconds":13.126,"score":11,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-dda578e283a67e43cb93.json"},"rank":17},{"modelId":"meta/muse-wonder","chainId":"latch-five-levels","family":"latch","score":67,"cognitiveScore":67,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":8.5199999616,"score":20.6,"turns":27,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/ec2ab747fc0a6a01ee2e.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/a43cf9c7ff0d3acf6866.svg","durationSeconds":10.74,"score":9,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/baa061c9835ffade4dd8.svg","durationSeconds":12.82,"score":18.8,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/f55abe0d4de61cb11ad6.svg","durationSeconds":11.6,"score":17.8,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/df03bd090d7b1fed1112.svg","durationSeconds":11.52,"score":20.6,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-5121184f2ce39d916a4a.json"},"rank":17},{"modelId":"union-alpha","chainId":"foresight-self-improvement","family":"foresight","score":78.7,"cognitiveScore":78.7,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/e1384c325d6bc9b831b3.png","previewScore":{"t":12.46,"score":19,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-17","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/bb9c8a135479e3d62622.svg","durationSeconds":17.416,"score":27,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/3633163a02b3c7ecb3fb.svg","durationSeconds":11.556,"score":12,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/92ca80dbfcc2a7ec252f.svg","durationSeconds":13.96,"score":19,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-5af40ae66c7720fe48a2.json"},"rank":1},{"modelId":"union-alpha","chainId":"latch-five-levels","family":"latch","score":36.5,"cognitiveScore":36.5,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":10.079999976000002,"score":19.4,"turns":33,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/980f860d870bcd8afe3c.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2fc96d78a65ae6e09baa.svg","durationSeconds":11.34,"score":14.2,"outcome":"cleared","finished":true,"beaconsLeft":4},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/802cb35f861329977e5b.svg","durationSeconds":12.38,"score":9.2,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/f005de52922a896ba158.svg","durationSeconds":13.08,"score":19.4,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-917c05f2cfcdece3e828.json"},"rank":30},{"modelId":"reflection/beam-1.0-501b-a23b","chainId":"foresight-self-improvement","family":"foresight","score":49.4,"cognitiveScore":49.4,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/af70d0ade74f110c14a6.png","previewScore":{"t":12.508,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-16","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/c0ff8789eda9f2e0ee8b.svg","durationSeconds":15.286,"score":18,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/c6bca933bfb3acef2688.svg","durationSeconds":10.85,"score":9,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/a36d9b59f09b79d1591c.svg","durationSeconds":13.608,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-2d77578098a5e6162ab7.json"},"rank":18},{"modelId":"reflection/beam-1.0-501b-a23b","chainId":"latch-five-levels","family":"latch","score":37.3,"cognitiveScore":37.3,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/36ad0618b72151fab422.png","previewScore":{"t":9.5600000496,"score":5,"turns":30,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/a3639a86d58d88070c9d.svg","durationSeconds":9.96,"score":9.6,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/a68017d592c883a7fab0.svg","durationSeconds":10.04,"score":11.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/37af997017aea17616a0.svg","durationSeconds":11.86,"score":3.6,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/15f543f9984de489f5ad.svg","durationSeconds":10.04,"score":5,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-9b6f77fd6ec70e357580.json"},"rank":29},{"modelId":"deepseek/deepseek-v4.1-flash-1017","chainId":"foresight-self-improvement","family":"foresight","score":52.9,"cognitiveScore":52.9,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/60bc96c1bc36bdb17aed.png","previewScore":{"t":11.162,"score":10,"turns":15,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/86f66ce837eb2ce7a164.svg","durationSeconds":17.69,"score":26,"outcome":"cleared"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/775846d9fea8b8aeb232.svg","durationSeconds":10.808,"score":7,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/48d65bf8dfedc2afbfa1.svg","durationSeconds":12.262,"score":10,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-7ba9ad3db65c5c888035.json"},"rank":15},{"modelId":"deepseek/deepseek-v4.1-flash-1017","chainId":"latch-five-levels","family":"latch","score":73.9,"cognitiveScore":73.9,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":9.299999988,"score":20,"turns":30,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-18","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/a7e64422840cadd13b3f.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2ef1b80d5b5e63b85151.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/5672e5b2af3e430eeea1.svg","durationSeconds":10.82,"score":16.4,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/6c1b380c8940916654ee.svg","durationSeconds":12.3,"score":20,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-bbc0f8d7a0c94571e93f.json"},"rank":13},{"modelId":"minimax/minimax-m3.1-flash","chainId":"foresight-self-improvement","family":"foresight","score":30.7,"cognitiveScore":30.7,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/8a19ed22d800d9ce6ef1.png","previewScore":{"t":13.49,"score":9,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-19","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/02ef6d0c65877484bb8e.svg","durationSeconds":8.69,"score":6,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/6e56cd4357115c5f6f20.svg","durationSeconds":12.22,"score":5,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/c8640dd7802abaaa65a4.svg","durationSeconds":14.59,"score":9,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-2f8a3e83241bd0e6c041.json"},"rank":28},{"modelId":"minimax/minimax-m3.1-flash","chainId":"latch-five-levels","family":"latch","score":21.9,"cognitiveScore":21.9,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/65d86f8a9e845f159aa0.png","previewScore":{"t":8.0000000304,"score":17.2,"turns":24,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-19","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/c928f1f56ec2e4d3793e.svg","durationSeconds":7.7,"score":8.6,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/f6bd5eed144397736e97.svg","durationSeconds":11.34,"score":5,"outcome":"out-of-turns"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/541f70fdd6e15b190470.svg","durationSeconds":12.12,"score":4.4,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/6b1a11a3846e17e75014.svg","durationSeconds":8.48,"score":17.2,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-c3f496041a3b5958892f.json"},"rank":38},{"modelId":"xiaomi/mimo-v2.6-pro","chainId":"foresight-self-improvement","family":"foresight","score":54.6,"cognitiveScore":54.6,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/34c1ae9048e14554dc6c.png","previewScore":{"t":12.46,"score":19,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-22","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/a640eba5104ccb764582.svg","durationSeconds":7.826,"score":4,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/3dc74a3d56cca9f1be91.svg","durationSeconds":10.626,"score":9,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/f5fb11244b08a7936e14.svg","durationSeconds":13.96,"score":19,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-24e06fbbf00d16d48370.json"},"rank":13},{"modelId":"xiaomi/mimo-v2.6-pro","chainId":"latch-five-levels","family":"latch","score":52.6,"cognitiveScore":52.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/e9bb91c4733d164816a3.png","previewScore":{"t":8.7800000046,"score":20.4,"turns":28,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-22","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/9884a7009943786c0da6.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/d406f3887e0dd2ca5e09.svg","durationSeconds":7.18,"score":9,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/5c86e0e3e917e31567c6.svg","durationSeconds":11.34,"score":17.2,"outcome":"cleared","finished":true,"beaconsLeft":1},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/da10cffb10cc17e52506.svg","durationSeconds":11.34,"score":16,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/702b5563307f2e7175a3.svg","durationSeconds":11.78,"score":20.4,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-739464991f953c9f9ad4.json"},"rank":22},{"modelId":"moonshotai/kimi-k3.1-pro","chainId":"foresight-self-improvement","family":"foresight","score":61.1,"cognitiveScore":61.1,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/a371dff8fadeab54061f.png","previewScore":{"t":12.734,"score":15,"turns":20,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-23","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/1c2c80f31ace24ba5d4c.svg","durationSeconds":12.146,"score":15,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/e073d510b290e409ccc7.svg","durationSeconds":11.556,"score":12,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/1cdba352ae5738b1f909.svg","durationSeconds":14.234,"score":15,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-8592bdf110866437fc27.json"},"rank":10},{"modelId":"moonshotai/kimi-k3.1-pro","chainId":"latch-five-levels","family":"latch","score":81.8,"cognitiveScore":81.8,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/6b9f3fcdfaa4753ef2cf.png","previewScore":{"t":9.56000004,"score":19.8,"turns":31,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-23","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/ec2ab747fc0a6a01ee2e.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/ed794aca2497416cfd1f.svg","durationSeconds":9.7,"score":9.8,"outcome":"cleared"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/000a94db4d339fd10fdd.svg","durationSeconds":12.3,"score":19.2,"outcome":"cleared"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/1bc1e19b39450888020f.svg","durationSeconds":8.48,"score":18.2,"outcome":"cleared","finished":true,"beaconsLeft":3},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/b51a8c4bfa5f126d03d9.svg","durationSeconds":12.56,"score":19.8,"outcome":"cleared"}],"detail":"/data/v1/benchmark-detail/replay-8e71c538bf487054c5c8.json"},"rank":8},{"modelId":"iceman-contributor","chainId":"latch-five-levels","family":"latch","score":3.6,"cognitiveScore":3.6,"coverage":{"completed":5,"selected":5},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn.","summary":"5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Hold the door","turns":12},{"label":"Level 2","name":"Crossfeed","turns":24},{"label":"Level 3","name":"Relay vault","turns":36},{"label":"Level 4","name":"Passing bay","turns":40},{"label":"Level 5","name":"Second thoughts","turns":40}]},"stages":[{"id":"hold-the-door","name":"Hold the door","label":"Level 1","settings":{"rows":["#######","#..#..#","#.....#","#..#..#","#######"],"steps":12,"robots":[[1,2],[1,3]],"cores":[{"at":[2,2],"goal":[5,2],"points":2}],"doors":[{"at":[3,2],"plate":[2,3]}],"beacons":[[2,3],[1,1]]}},{"id":"crossfeed","name":"Crossfeed","label":"Level 2","settings":{"rows":["#########","#...#...#","#.......#","#...#...#","#.......#","#...#...#","#########"],"steps":24,"robots":[[1,1],[1,5]],"cores":[{"at":[3,2],"goal":[5,1],"points":2},{"at":[3,4],"goal":[7,4],"points":4}],"doors":[{"at":[4,2],"plate":[2,4]},{"at":[4,4],"plate":[5,2]}],"beacons":[[7,4],[1,2],[2,5]]}},{"id":"relay-vault","name":"Relay vault","label":"Level 3","settings":{"rows":["###########","#...#...###","#.......###","#...#...#.#","#...#.....#","#...#...#.#","#.......###","#...#...###","###########"],"steps":36,"robots":[[1,1],[1,7]],"cores":[{"at":[3,2],"goal":[6,2],"points":2},{"at":[3,6],"goal":[6,6],"points":4},{"at":[7,4],"goal":[9,4],"points":8}],"doors":[{"at":[4,2],"plate":[2,5]},{"at":[4,6],"plate":[6,2]},{"at":[8,4],"plate":[6,6]}],"beacons":[[9,3],[9,5],[7,2],[9,4]]}},{"id":"passing-bay","name":"Passing bay","label":"Level 4","settings":{"rows":["############","#..........#","#######....#","#..........#","#.#####....#","#.....#....#","############"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":8},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[],"beacons":[[1,1],[5,5],[4,5],[7,1]]}},{"id":"second-thoughts","name":"Second thoughts","label":"Level 5","settings":{"rows":["###########","#...##....#","#...##....#","#.........#","#.####...##","########.##","###########"],"steps":40,"robots":[[3,3],[1,3]],"cores":[{"at":[4,3],"goal":[5,3],"points":4},{"at":[2,3],"goal":[9,3],"points":6},{"at":[8,4],"goal":[8,5],"points":2}],"doors":[{"at":[6,3],"plate":[8,4]}],"beacons":[[2,3],[2,2],[1,2],[8,3],[5,3],[8,4]]}}],"preview":"/shared/v1/game-replays/626b5325b2a4802cd358.png","previewScore":{"t":8.2599999866,"score":5,"turns":25,"limit":40},"previewLabel":"End · Level 5/5","date":"2026-10-25","replay":{"stages":[{"label":"Level 1 of 5 · Hold the door","available":true,"src":"/shared/v1/game-replays/b609b513ee5e2b6306d0.svg","durationSeconds":6.06,"score":5.2,"outcome":"cleared"},{"label":"Level 2 of 5 · Crossfeed","available":true,"src":"/shared/v1/game-replays/c5d686209c54d21aa490.svg","durationSeconds":7.44,"score":2.8,"outcome":"ended-early"},{"label":"Level 3 of 5 · Relay vault","available":true,"src":"/shared/v1/game-replays/2167ecd1d4b569dd0e14.svg","durationSeconds":10.04,"score":2.2,"outcome":"ended-early"},{"label":"Level 4 of 5 · Passing bay","available":true,"src":"/shared/v1/game-replays/3858ee1415c37a707d11.svg","durationSeconds":8.74,"score":4,"outcome":"ended-early"},{"label":"Level 5 of 5 · Second thoughts","available":true,"src":"/shared/v1/game-replays/ff28b33e2be47f2cb925.svg","durationSeconds":8.74,"score":5,"outcome":"ended-early"}],"detail":"/data/v1/benchmark-detail/replay-132c7d519b3984fb69a6.json"},"rank":44},{"modelId":"iceman-contributor","chainId":"foresight-self-improvement","family":"foresight","score":3.2,"cognitiveScore":3.2,"coverage":{"completed":3,"selected":3},"inContextLearning":true,"earlierTask":false,"task":{"rules":"Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life.","summary":"3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.","stages":[{"label":"Level 1","name":"Second Thoughts","turns":28},{"label":"Level 2","name":"Pressure Pot","turns":18},{"label":"Level 3","name":"Hide or Flight","turns":20}]},"stages":[{"id":"custom-eddab0e3-c4af-4138-b47f-a06025b67843","name":"Second Thoughts","label":"Level 1","settings":{"rows":["..#......",".#######.",".........","##.###.#.","#...#....","##.###.#.",".........","#.#.#.##.","........."],"steps":28,"starts":{"player":[4,8,"LEFT"],"ghosts":[[7,6,"LEFT"],[6,0,"LEFT"],[4,6,"LEFT"],[5,4,"RIGHT"]]},"tunnelRow":null,"lives":1}},{"id":"foresight-17","name":"Pressure Pot","label":"Level 2","settings":{"rows":[".............",".#####.#####.",".#.........#.",".#.###.###.#.","...#.....#...",".###.#.#.###.",".............",".###.###.###.","...#.....#...",".#.###.###.#.",".##.......##.",".##..#.#.###.","..#.......#.."],"steps":18,"starts":{"player":[4,12,"RIGHT"],"ghosts":[[8,8,"LEFT"],[2,6,"RIGHT"],[4,6,"DOWN"],[8,5,"UP"]]},"lives":1}},{"id":"original-15","name":"Hide or Flight","label":"Level 3","settings":{"rows":[".#...#...","...#...#.",".#####...",".#.#.####","......##.","#..##..#.",".......#.","##.###.#.","........."],"steps":20,"starts":{"player":[4,8,"LEFT"],"ghosts":[[4,3,"DOWN"],[4,6,"LEFT"],[2,3,"DOWN"],[0,6,"RIGHT"]]},"tunnelRow":null,"lives":1}}],"preview":"/shared/v1/game-replays/76c4266be1ba0b6aceed.png","previewScore":{"t":8.894,"score":1,"turns":7,"limit":20},"previewLabel":"End · Level 3/3","date":"2026-10-25","replay":{"stages":[{"label":"Level 1 of 3 · Second Thoughts","available":true,"src":"/shared/v1/game-replays/3ca33e3e1d2da2a11830.svg","durationSeconds":8.49,"score":0,"outcome":"out-of-lives"},{"label":"Level 2 of 3 · Pressure Pot","available":true,"src":"/shared/v1/game-replays/573e9d42d176d42db537.svg","durationSeconds":10.492,"score":3,"outcome":"out-of-lives"},{"label":"Level 3 of 3 · Hide or Flight","available":true,"src":"/shared/v1/game-replays/9fd25dde72864b324c3b.svg","durationSeconds":9.994,"score":1,"outcome":"out-of-lives"}],"detail":"/data/v1/benchmark-detail/replay-ac2d7d7835b6f3af56d5.json"},"rank":47}],"presentation":{"maxVisibleColumns":null,"minColumnWidth":196},"references":{"newtons-mirror":{"src":"/benchmarks/media/13cad1a6715c9b84.png","width":3200,"height":2400,"label":"Beauty render","display":{"src":"/benchmarks/media/5b130a89c0822a26.webp","width":1600,"height":1200}},"ground-truth":{"src":"/benchmarks/media/9254bfe89a90f7ce.png","width":1080,"height":1920,"label":"Beauty render"},"mona-lisa":{"src":"/benchmarks/media/803675074021c0c7.png","width":606,"height":791,"label":"Mona Lisa at the Louvre"}},"comparison":{"sceneOrder":["newtons-mirror","selfie","mona-lisa","beach","ground-truth","raccoon","foresight-self-improvement","latch-five-levels","pelican"],"observations":{"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"cat-jam\",\"\"]]":{"groupId":"mona-lisa","text":"Cat Jam Free scores higher on Mona Lisa, with its largest category advantage in space & placement."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"anthropic/claude-fable-5.1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"anthropic/claude-haiku-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"anthropic/claude-opus-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"anthropic/claude-sonnet-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"deepseek/deepseek-v4-pro-0813\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"deepseek/deepseek-v4.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in materials and scale."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in materials and scale."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in materials and scale."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ox Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Beam 1.0 501B A23B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"reflection/beam-1.0-501b-a23b\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"anthropic/claude-fable-5.1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"anthropic/claude-haiku-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"anthropic/claude-opus-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"anthropic/claude-sonnet-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"deepseek/deepseek-v4-pro-0813\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"cat-jam\",\"\"],[\"deepseek/deepseek-v4.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"cat-jam\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"cat-jam\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"beach","text":"Cat Jam Free scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"cat-jam\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"raccoon","text":"Cat Jam Free scores higher on Raccoon, with its largest category advantage in light & reflections."},"[[\"cat-jam\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"cat-jam\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"ground-truth","text":"Cat Jam Free scores higher on Ground Truth."},"[[\"cat-jam\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"raccoon","text":"Cat Jam Free scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"cat-jam\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in materials and scale."},"[[\"cat-jam\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"cat-jam\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"cat-jam\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"cat-jam\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"cat-jam\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Cat Jam Free scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"cat-jam\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"Cat Jam Free scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"cat-jam\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"ground-truth","text":"Cat Jam Free scores higher on Ground Truth."},"[[\"cat-jam\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"ground-truth","text":"MiMo-V2.6-Flash scores higher on Ground Truth."},"[[\"cat-jam\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"cat-jam\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"cat-jam\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"raccoon","text":"Cat Jam Free scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"cat-jam\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Cat Jam Free scores higher on Ground Truth."},"[[\"cat-jam\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"cat-jam\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"cat-jam\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"Cat Jam Free scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"cat-jam\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"raccoon","text":"Cat Jam Free scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"cat-jam\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Cat Jam Free scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"cat-jam\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in space & placement."},"[[\"cat-jam\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"cat-jam\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"anthropic/claude-haiku-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"anthropic/claude-opus-5.5\",\"\"]]":{"groupId":"ground-truth","text":"Claude Opus 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"anthropic/claude-sonnet-5.5\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"deepseek/deepseek-v4-pro-0813\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"deepseek/deepseek-v4.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Claude Fable 5.1 scores higher on Ground Truth."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"beach","text":"Claude Fable 5.1 scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"raccoon","text":"Gemini 4 Argon scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"raccoon","text":"GPT-6.1 Sol scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"mona-lisa","text":"Claude Fable 5.1 scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-fable-5.1\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Fable 5.1 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"anthropic/claude-opus-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"anthropic/claude-sonnet-5.5\",\"\"]]":{"groupId":"selfie","text":"Claude Sonnet 5.5 scores higher on Selfie, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"deepseek/deepseek-v4-pro-0813\",\"\"]]":{"groupId":"beach","text":"Claude Haiku 5.5 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"deepseek/deepseek-v4.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"beach","text":"Claude Haiku 5.5 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"mona-lisa","text":"Claude Haiku 5.5 scores higher on Mona Lisa, with its largest category advantage in light & reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"raccoon","text":"Kimi K3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"raccoon","text":"Kimi K3.1 Pro scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Claude Haiku 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"beach","text":"Claude Haiku 5.5 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Claude Haiku 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"raccoon","text":"Qwen3.8-27B scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"raccoon","text":"Qwen3.8-Max-0902 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Claude Haiku 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-haiku-5.5\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"anthropic/claude-sonnet-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"deepseek/deepseek-v4-pro-0813\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"deepseek/deepseek-v4.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Claude Opus 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"ground-truth","text":"Claude Opus 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-opus-5.5\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Opus 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"deepseek/deepseek-v4-pro-0813\",\"\"]]":{"groupId":"mona-lisa","text":"Claude Sonnet 5.5 scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"deepseek/deepseek-v4.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"beach","text":"Claude Sonnet 5.5 scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Claude Sonnet 5.5 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"anthropic/claude-sonnet-5.5\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Claude Sonnet 5.5 scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"deepseek/deepseek-v4.1-flash\",\"\"]]":{"groupId":"selfie","text":"DeepSeek-V4.1-Flash scores higher on Selfie, with its largest category advantage in artistry & character."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"mona-lisa","text":"DeepSeek-V4.1-Flash-1017 scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"beach","text":"GPT-5.6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"beach","text":"GPT-6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"beach","text":"Hy4 preview scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"beach","text":"Kimi K3 scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"beach","text":"Kimi K3.1 Pro scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"beach","text":"Ling 3.1 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"mona-lisa","text":"MiMo-V2.6-Pro scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.8-27B scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Max-0902 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4-Pro-0813 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4-pro-0813\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"deepseek/deepseek-v4.1-flash-1017\",\"\"]]":{"groupId":"ground-truth","text":"DeepSeek-V4.1-Flash-1017 scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6.1 Sol scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"mona-lisa","text":"DeepSeek-V4.1-Flash scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Max-0902 scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Union Alpha scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"fledge-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"beach","text":"Gemini 4 Argon scores higher on Beach, with its largest category advantage in space & placement."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"beach","text":"GPT-6.1 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"ground-truth","text":"DeepSeek-V4.1-Flash-1017 scores higher on Ground Truth."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"mona-lisa","text":"DeepSeek-V4.1-Flash-1017 scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"mona-lisa","text":"DeepSeek-V4.1-Flash-1017 scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Max-0902 scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"DeepSeek-V4.1-Flash-1017 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"deepseek/deepseek-v4.1-flash-1017\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in materials and scale."},"[[\"fledge-alpha\",\"\"],[\"google/gemini-3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"fledge-alpha\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"fledge-alpha\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"fledge-alpha\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"ground-truth","text":"Fledge Alpha scores higher on Ground Truth."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Fledge Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"fledge-alpha\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"ground-truth","text":"Fledge Alpha scores higher on Ground Truth."},"[[\"fledge-alpha\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"fledge-alpha\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"ground-truth","text":"Fledge Alpha scores higher on Ground Truth."},"[[\"fledge-alpha\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"fledge-alpha\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"fledge-alpha\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Fledge Alpha scores higher on Ground Truth."},"[[\"fledge-alpha\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ox Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"fledge-alpha\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"raccoon","text":"Fledge Alpha scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"fledge-alpha\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"ground-truth","text":"Fledge Alpha scores higher on Ground Truth."},"[[\"fledge-alpha\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"fledge-alpha\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"fledge-alpha\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"google/gemini-4-argon\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-3.8-flash\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"raccoon","text":"Gemini 3.8 Flash scores higher on Raccoon, with its largest category advantage in artistry & character."},"[[\"google/gemini-3.8-flash\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-5.6 Sol scores higher on Mona Lisa, with its largest category advantage in visual integrity."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6 Luna scores higher on Mona Lisa, with its largest category advantage in visual integrity."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6 Sol scores higher on Mona Lisa, with its largest category advantage in visual integrity."},"[[\"google/gemini-3.8-flash\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"mona-lisa","text":"Kimi K3 scores higher on Mona Lisa, with its largest category advantage in visual integrity."},"[[\"google/gemini-3.8-flash\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-3.8-flash\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-3.8-flash\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"google/gemini-3.8-flash\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"ground-truth","text":"Gemini 3.8 Flash scores higher on Ground Truth."},"[[\"google/gemini-3.8-flash\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Gemini 3.8 Flash scores higher on Ground Truth."},"[[\"google/gemini-3.8-flash\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"google/gemini-3.8-flash\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"beach","text":"Gemini 3.8 Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-3.8-flash\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"google/gemini-3.8-flash\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-3.8-flash\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Gemini 3.8 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemini-3.8-flash\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"google/gemma-4-26b-a4b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"mona-lisa","text":"Gemini 4 Argon scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"google/gemini-4-argon\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemini-4-argon\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemini 4 Argon scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"google/gemma-4-31b-it\",\"\"]]":{"groupId":"mona-lisa","text":"Gemma 4 31B scores higher on Mona Lisa, with its largest category advantage in space & placement."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Glyph Cluster scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"Gemma 4 26B A4B scores higher on Raccoon, with its largest category advantage in artistry & character."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"ground-truth","text":"Gemma 4 26B A4B scores higher on Ground Truth."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"beach","text":"MiMo-V2.6-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"raccoon","text":"Gemma 4 26B A4B scores higher on Raccoon, with its largest category advantage in visual integrity."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"Gemma 4 26B A4B scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-27B scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-35B-A3B scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-26b-a4b-it\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"z-ai/glm-5.3\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemma 4 31B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"ground-truth","text":"GPT-5.5 scores higher on Ground Truth."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"raccoon","text":"GPT-5.6 Luna scores higher on Raccoon, with its largest category advantage in light & reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"beach","text":"GPT-6 Luna scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"beach","text":"GPT-6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemma 4 31B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-31b-it\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"Gemma 4 31B scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemma 4 31B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-31b-it\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"beach","text":"MiMo-V2.6-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"google/gemma-4-31b-it\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"google/gemma-4-31b-it\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Gemma 4 31B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"google/gemma-4-31b-it\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"google/gemma-4-31b-it\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"google/gemma-4-31b-it\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"Gemma 4 31B scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-27B scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-35B-A3B scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"google/gemma-4-31b-it\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"google/gemma-4-31b-it\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"google/gemma-4-31b-it\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"google/gemma-4-31b-it\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"glyph-cluster\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-5.6 Sol scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6 Sol scores higher on Mona Lisa, with its largest category advantage in prompt & likeness."},"[[\"z-ai/glm-5.3\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"z-ai/glm-5.3\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"z-ai/glm-5.3\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"missingno\",\"\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"z-ai/glm-5.3\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in materials and scale."},"[[\"z-ai/glm-5.3\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"beach","text":"GLM-5.3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"z-ai/glm-5.3\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GLM-5.3 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"z-ai/glm-5.3\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"z-ai/glm-5.3\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"ground-truth","text":"GLM-5.3 scores higher on Ground Truth."},"[[\"z-ai/glm-5.3\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-5.5\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Glyph Cluster scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Glyph Cluster scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"ground-truth","text":"Glyph Cluster scores higher on Ground Truth."},"[[\"glyph-cluster\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"ground-truth","text":"MiMo-V2.6-Flash scores higher on Ground Truth."},"[[\"glyph-cluster\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Spark 1.3 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"glyph-cluster\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"glyph-cluster\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Glyph Cluster scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"glyph-cluster\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"ground-truth","text":"Glyph Cluster scores higher on Ground Truth."},"[[\"glyph-cluster\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.6-27B scores higher on Ground Truth."},"[[\"glyph-cluster\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Glyph Cluster scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"glyph-cluster\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"glyph-cluster\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"openai/gpt-5.6-luna\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-5.6 Luna scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"beach","text":"GPT-5.6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"raccoon","text":"GPT-5.6 Terra scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"openai/gpt-5.5\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6 Luna scores higher on Mona Lisa, with its largest category advantage in space & placement."},"[[\"openai/gpt-5.5\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"beach","text":"GPT-6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.5\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"beach","text":"Kimi K3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.5\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"ground-truth","text":"GPT-5.5 scores higher on Ground Truth."},"[[\"openai/gpt-5.5\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"beach","text":"MiMo-V2.6-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"raccoon","text":"MiniMax M3 scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"openai/gpt-5.5\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"openai/gpt-5.5\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.5\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GPT-5.5 scores higher on Ground Truth."},"[[\"openai/gpt-5.5\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"openai/gpt-5.5\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-5.5\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.5\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-27B scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.5 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.5\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.5\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.5\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"openai/gpt-5.6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"selfie","text":"GPT-5.6 Luna scores higher on Selfie, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"beach","text":"GPT-6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"GPT-5.6 Luna scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-5.6 Luna scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"raccoon","text":"GPT-5.6 Luna scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GPT-5.6 Luna scores higher on Ground Truth."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"GPT-5.6 Luna scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"openai/gpt-5.6-luna\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"openai/gpt-5.6-terra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"selfie","text":"GPT-5.6 Sol scores higher on Selfie, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-5.6 Sol scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"raccoon","text":"Kimi K3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"beach","text":"GPT-5.6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-5.6 Sol scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GPT-5.6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-sol\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"openai/gpt-6-astra\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"selfie","text":"GPT-5.6 Terra scores higher on Selfie, with its largest category advantage in prompt & likeness."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"GPT-5.6 Terra scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-5.6 Terra scores higher on Mona Lisa, with its largest category advantage in visual integrity."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"missingno\",\"\"]]":{"groupId":"ground-truth","text":"MISSINGNO. scores higher on Ground Truth."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GPT-5.6 Terra scores higher on Ground Truth."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"GPT-5.6 Terra scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"GPT-5.6 Terra scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-5.6-terra\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"openai/gpt-6-luna\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-astra\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"openai/gpt-6-sol\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-luna\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"GPT-6 Luna scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"openai/gpt-6-luna\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-luna\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-luna\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6 Luna scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-6-luna\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"missingno\",\"\"]]":{"groupId":"ground-truth","text":"MISSINGNO. scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"ground-truth","text":"Mistral Large 4 scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6 Luna scores higher on Mona Lisa, with its largest category advantage in space & placement."},"[[\"openai/gpt-6-luna\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-luna\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Luna scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-6-luna\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"GPT-6 Luna scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-6-luna\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-luna\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Luna scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-luna\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"openai/gpt-6-luna\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-luna\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"openai/gpt-6-luna\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"openai/gpt-6.1-sol\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"raccoon","text":"GPT-6 Sol scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-6-sol\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-6-sol\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6 Sol scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"openai/gpt-6-sol\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"missingno\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-6-sol\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-6-sol\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-6-sol\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6 Sol scores higher on Ground Truth."},"[[\"openai/gpt-6-sol\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"openai/gpt-6-sol\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"beach","text":"GPT-6 Sol scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-6-sol\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"openai/gpt-6-sol\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6-sol\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"GPT-6 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6-sol\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"x-ai/grok-4.7\",\"\"]]":{"groupId":"mona-lisa","text":"GPT-6.1 Sol scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"GPT-6.1 Sol scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"openai/gpt-6.1-sol\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"ground-truth","text":"GPT-6.1 Sol scores higher on Ground Truth."},"[[\"x-ai/grok-4.7\",\"\"],[\"tencent/hy4-preview\",\"\"]]":{"groupId":"beach","text":"Hy4 preview scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"x-ai/grok-4.7\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"x-ai/grok-4.7\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"beach","text":"Kimi K3 scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"x-ai/grok-4.7\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"mona-lisa","text":"Kimi K3.1 Pro scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"x-ai/grok-4.7\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"beach","text":"Ling 3.1 Flash scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"x-ai/grok-4.7\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"ground-truth","text":"MiMo-V2.6-Pro scores higher on Ground Truth."},"[[\"x-ai/grok-4.7\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"x-ai/grok-4.7\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"MiniMax M3.1 Flash scores higher on Ground Truth."},"[[\"x-ai/grok-4.7\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"x-ai/grok-4.7\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"x-ai/grok-4.7\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"x-ai/grok-4.7\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"x-ai/grok-4.7\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-27B scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"x-ai/grok-4.7\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"x-ai/grok-4.7\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Max-0902 scores higher on Ground Truth."},"[[\"x-ai/grok-4.7\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"x-ai/grok-4.7\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Grok 4.7 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"x-ai/grok-4.7\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Union Alpha scores higher on Ground Truth."},"[[\"tencent/hy4-preview\",\"\"],[\"iceman-contributor\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"raccoon","text":"Kimi K3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"tencent/hy4-preview\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"raccoon","text":"Kimi K3.1 Pro scores higher on Raccoon, with its largest category advantage in artistry & character."},"[[\"tencent/hy4-preview\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Hy4 preview scores higher on Ground Truth."},"[[\"tencent/hy4-preview\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"mona-lisa","text":"MiMo-V2.6-Pro scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"tencent/hy4-preview\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"tencent/hy4-preview\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"tencent/hy4-preview\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.8-27B scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"tencent/hy4-preview\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.8-Max-0902 scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"tencent/hy4-preview\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Hy4 preview scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"tencent/hy4-preview\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"moonshotai/kimi-k3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Kolibri-1 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Iceman Contributor scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"iceman-contributor\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"iceman-contributor\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Spark 1.3 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"iceman-contributor\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Spark Open scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Iceman Contributor scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"iceman-contributor\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ox Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"iceman-contributor\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Iceman Contributor scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"iceman-contributor\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.6-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Iceman Contributor scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"iceman-contributor\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"iceman-contributor\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"moonshotai/kimi-k3.1-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"beach","text":"Kimi K3 scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"moonshotai/kimi-k3\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"beach","text":"Kimi K3 scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"moonshotai/kimi-k3\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"raccoon","text":"Kimi K3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"moonshotai/kimi-k3\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"mona-lisa","text":"Kimi K3 scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"moonshotai/kimi-k3\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"moonshotai/kimi-k3\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"moonshotai/kimi-k3\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Kimi K3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"Aleph-Alpha/Kolibri-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Kimi K3.1 Pro scores higher on Ground Truth."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"raccoon","text":"Kimi K3.1 Pro scores higher on Raccoon, with its largest category advantage in artistry & character."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"mona-lisa","text":"Kimi K3.1 Pro scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"Kimi K3.1 Pro scores higher on Ground Truth."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"raccoon","text":"Qwen3.8-27B scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Max-0902 scores higher on Ground Truth."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Kimi K3.1 Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"moonshotai/kimi-k3.1-pro\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"inclusionai/ling-3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"ground-truth","text":"Kolibri-1 scores higher on Ground Truth."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"beach","text":"MiMo-V2.6-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"raccoon","text":"MiniMax M3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"beach","text":"Nemotron 3 Ultra scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ox Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"ground-truth","text":"Kolibri-1 scores higher on Ground Truth."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-27B scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-35B-A3B scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"Aleph-Alpha/Kolibri-1\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"microsoft/mai-thinking-1\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"mona-lisa","text":"MiMo-V2.6-Pro scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"ground-truth","text":"MiniMax M3.1 Flash scores higher on Ground Truth."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"ground-truth","text":"Pixel Canary scores higher on Ground Truth."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.8-27B scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Max-0902 scores higher on Ground Truth."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Ling 3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"inclusionai/ling-3.1-flash\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"xiaomi/mimo-v2.6-flash\",\"\"]]":{"groupId":"ground-truth","text":"MiMo-V2.6-Flash scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark 1.3 scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"ground-truth","text":"Nemotron 3 Ultra scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.5-9B scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.6-27B scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.6-35B-A3B scores higher on Ground Truth."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"microsoft/mai-thinking-1\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"xiaomi/mimo-v2.6-pro\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"beach","text":"MiMo-V2.6-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Spark Open scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"mona-lisa","text":"MiMo-V2.6-Flash scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"MiMo-V2.6-Flash scores higher on Ground Truth."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ox Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"beach","text":"MiMo-V2.6-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"ground-truth","text":"MiMo-V2.6-Flash scores higher on Ground Truth."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"ground-truth","text":"MiMo-V2.6-Flash scores higher on Ground Truth."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-flash\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"minimax/minimax-m3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"raccoon","text":"Qwen3.8-27B scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Max-0902 scores higher on Ground Truth."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"MiMo-V2.6-Pro scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"xiaomi/mimo-v2.6-pro\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3\",\"\"],[\"minimax/minimax-m3.1-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3\",\"\"],[\"missingno\",\"\"]]":{"groupId":"ground-truth","text":"MISSINGNO. scores higher on Ground Truth."},"[[\"minimax/minimax-m3\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"ground-truth","text":"Mistral Large 4 scores higher on Ground Truth."},"[[\"minimax/minimax-m3\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"minimax/minimax-m3\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"minimax/minimax-m3\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"ground-truth","text":"Pixel Canary scores higher on Ground Truth."},"[[\"minimax/minimax-m3\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"MiniMax M3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"minimax/minimax-m3\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"minimax/minimax-m3\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"minimax/minimax-m3\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"minimax/minimax-m3\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"missingno\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"mona-lisa","text":"MiniMax M3.1 Flash scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Max-0902 scores higher on Ground Truth."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"MiniMax M3.1 Flash scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"minimax/minimax-m3.1-flash\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"missingno\",\"\"],[\"mistralai/mistral-large-4-0\",\"\"]]":{"groupId":"beach","text":"Mistral Large 4 scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"missingno\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"missingno\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"MISSINGNO. scores higher on Ground Truth."},"[[\"missingno\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"beach","text":"Ox Alpha scores higher on Beach, with its largest category advantage in artistry & character."},"[[\"missingno\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"mona-lisa","text":"Pixel Canary scores higher on Mona Lisa, with its largest category advantage in prompt & likeness."},"[[\"missingno\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"missingno\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"missingno\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"missingno\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"MISSINGNO. scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"missingno\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"meta/muse-spark-1.3\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"mona-lisa","text":"Mistral Large 4 scores higher on Mona Lisa, with its largest category advantage in space & placement."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Mistral Large 4 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"mistralai/mistral-large-4-0\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-1.3\",\"\"],[\"meta/muse-spark-open\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"meta/muse-spark-1.3\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-spark-1.3\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"meta/muse-spark-1.3\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-1.3\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"meta/muse-spark-1.3\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-spark-1.3\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"meta/muse-spark-1.3\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in prompt & likeness."},"[[\"meta/muse-spark-1.3\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"raccoon","text":"Muse Spark 1.3 scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"meta/muse-spark-1.3\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-1.3\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"meta/muse-spark-1.3\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-1.3\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-1.3\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-1.3\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-open\",\"\"],[\"meta/muse-wonder\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-open\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"meta/muse-spark-open\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"meta/muse-spark-open\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ox Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-open\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"meta/muse-spark-open\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"meta/muse-spark-open\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"meta/muse-spark-open\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"meta/muse-spark-open\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-open\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"ground-truth","text":"Muse Spark Open scores higher on Ground Truth."},"[[\"meta/muse-spark-open\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-open\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-open\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-spark-open\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-wonder\",\"\"],[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-wonder\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"mona-lisa","text":"Muse Wonder scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"meta/muse-wonder\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-wonder\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-wonder\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-wonder\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-wonder\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-wonder\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"ground-truth","text":"Muse Wonder scores higher on Ground Truth."},"[[\"meta/muse-wonder\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-wonder\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"meta/muse-wonder\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-wonder\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Muse Wonder scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"meta/muse-wonder\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"owl-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Ox Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"beach","text":"Nemotron 3 Ultra scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.6-27B scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"raccoon","text":"Qwen3.6-35B-A3B scores higher on Raccoon, with its largest category advantage in anatomy & construction."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.8-Flash scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"nvidia/nemotron-3-ultra-550b-a55b\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"owl-alpha\",\"\"],[\"ox-alpha\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"owl-alpha\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"mona-lisa","text":"Pixel Canary scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"owl-alpha\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"owl-alpha\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"owl-alpha\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Owl Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"owl-alpha\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"owl-alpha\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Flash scores higher on Ground Truth."},"[[\"owl-alpha\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"owl-alpha\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"owl-alpha\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"ground-truth","text":"Space Bunny Alpha scores higher on Ground Truth."},"[[\"owl-alpha\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"ox-alpha\",\"\"],[\"pixel-canary\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"ox-alpha\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"ox-alpha\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"ox-alpha\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"ox-alpha\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"ox-alpha\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"ox-alpha\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"ox-alpha\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"ox-alpha\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"ground-truth","text":"Ox Alpha scores higher on Ground Truth."},"[[\"ox-alpha\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"pixel-canary\",\"\"],[\"qwen/qwen3.5-9b\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"pixel-canary\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"pixel-canary\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"pixel-canary\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"pixel-canary\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in camera and placement."},"[[\"pixel-canary\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"pixel-canary\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"pixel-canary\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Pixel Canary scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"pixel-canary\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"qwen/qwen3.6-27b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-27B scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"beach","text":"Qwen3.6-35B-A3B scores higher on Beach, with its largest category advantage in prompt & likeness."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"beach","text":"Qwen3.8-Flash scores higher on Beach, with its largest category advantage in anatomy & construction."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.5-9b\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-27b\",\"\"],[\"qwen/qwen3.6-35b-a3b\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.6-27B scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"qwen/qwen3.6-27b\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"qwen/qwen3.6-27b\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.8-Flash scores higher on Mona Lisa, with its largest category advantage in artistry & character."},"[[\"qwen/qwen3.6-27b\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-27b\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-27b\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-27b\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-35b-a3b\",\"\"],[\"qwen/qwen3.8-27B\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"qwen/qwen3.6-35b-a3b\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"mona-lisa","text":"Qwen3.8-Flash scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."},"[[\"qwen/qwen3.6-35b-a3b\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-35b-a3b\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-35b-a3b\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.6-35b-a3b\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-27B\",\"\"],[\"qwen/qwen3.8-flash\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-27B\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"ground-truth","text":"Qwen3.8-Max-0902 scores higher on Ground Truth."},"[[\"qwen/qwen3.8-27B\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"qwen/qwen3.8-27B\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-27B scores higher on Newton's Mirror, with its largest category advantage in state at 600 ms."},"[[\"qwen/qwen3.8-27B\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-flash\",\"\"],[\"qwen/qwen3.8-max-0902\",\"\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-flash\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-flash\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Space Bunny Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-flash\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-max-0902\",\"\"],[\"space-bunny-alpha\",\"selected-defaults\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-max-0902\",\"\"],[\"space-bunny-alpha\",\"from-20260929\"]]":{"groupId":"newtons-mirror","text":"Qwen3.8-Max-0902 scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"qwen/qwen3.8-max-0902\",\"\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"space-bunny-alpha\",\"selected-defaults\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"space-bunny-alpha\",\"from-20260929\"],[\"union-alpha\",\"\"]]":{"groupId":"newtons-mirror","text":"Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections."},"[[\"space-bunny-alpha\",\"before-20260929\"],[\"union-alpha\",\"\"]]":{"groupId":"mona-lisa","text":"Union Alpha scores higher on Mona Lisa, with its largest category advantage in anatomy & construction."}}},"stealthmarkVersion":"1.8.6","selectionScores":"/data/v1/benchmark-detail/selections-37660aaf0c78c500abe7.json"}