STEALTHMARK · HEAD TO HEAD

All comparisons ↗

GPT-6 Astra vs MISSINGNO.

GPT-6 Astra scores 32.5 points higher overall (70.5 vs 38.0). GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections.

See the difference ↘
All results ↗
Model guide ↗

Overall StealthMark

70.5/100
38.0/100

General Intelligence

64.0/100
33.6/100

Visual Intelligence

78.2/100
43.3/100

Newton's Mirror

Same task instructions

Physics, perspective and recursive reflections, tested together. Explore the test ↗

GPT-6 Astra62.7/100

GPT-6 Astra: Newton's Mirror SVG benchmark
Researcher observations

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.

Benchmark task

A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.

MISSINGNO.24.1/100

MISSINGNO.: Newton's Mirror SVG benchmark
Researcher observations

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.

Benchmark task

A spinning metal sphere and paint in a mirrored room test physical reasoning, perspective and consistency across repeated reflections.

Compare category scores
CategoryGPT-6 AstraMISSINGNO.
State at 600 ms25.0/2511.5/25
Camera and placement15.0/157.0/15
Reflections54.0/12012.0/120
Materials and scale10.0/105.5/10
Floor-damage reward0.0/100.0/10

Selfie

Different task instructions

GPT-6 Astra79.3/100

GPT-6 Astra: Selfie SVG benchmark
Researcher observations

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.

Benchmark task

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.

MISSINGNO.49.7/100

MISSINGNO.: Selfie SVG benchmark
Researcher observations

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.

Benchmark task

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.

Compare category scores
CategoryGPT-6 AstraMISSINGNO.
Task & likeness8.5/105.5/10
Anatomy & construction8.0/104.5/10
Space & placement7.5/106.0/10
Light & reflections7.0/105.5/10
Purposeful detail8.0/104.5/10
Artistry & character8.5/104.5/10
Materials & effects7.0/104.0/10
Visual integrity8.0/106.0/10

Mona Lisa

Different task instructions

GPT-6 Astra68.2/100

GPT-6 Astra: Mona Lisa SVG benchmark
Researcher observations

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.

Benchmark task

An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.

MISSINGNO.28.1/100

MISSINGNO.: Mona Lisa SVG benchmark
Researcher observations

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.

Benchmark task

An SVG interpretation of the Mona Lisa tests facial likeness, anatomy and the composition of a familiar portrait.

Compare category scores
CategoryGPT-6 AstraMISSINGNO.
Task & likeness7.0/101.5/10
Anatomy & construction6.5/102.5/10
Space & placement6.5/102.5/10
Light & reflections7.0/103.0/10
Purposeful detail6.5/102.0/10
Artistry & character7.0/103.5/10
Materials & effects7.0/102.0/10
Visual integrity7.5/106.0/10

Beach

Different task instructions

GPT-6 Astra81.8/100

GPT-6 Astra: Beach SVG benchmark
Researcher observations

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.

Benchmark task

A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.

MISSINGNO.43.9/100

MISSINGNO.: Beach SVG benchmark
Researcher observations

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.

Benchmark task

A busy shoreline at sunset tests scene composition, animal forms and the relationship between light and water reflections.

Compare category scores
CategoryGPT-6 AstraMISSINGNO.
Task & likeness9.0/107.0/10
Anatomy & construction7.5/104.0/10
Space & placement8.5/106.5/10
Light & reflections7.5/105.0/10
Purposeful detail8.5/103.0/10
Artistry & character9.0/103.0/10
Materials & effects7.5/105.0/10
Visual integrity8.0/103.5/10

Ground Truth

Same task instructions

GPT-6 Astra68.7/100

GPT-6 Astra: Ground Truth SVG benchmark
Benchmark task

A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.

MISSINGNO.44.1/100

MISSINGNO.: Ground Truth SVG benchmark
Benchmark task

A plank, computer mouse and blueberry in a steel room test physical stability, spatial relationships, materials and lighting.

Raccoon

Different task instructions

GPT-6 Astra78.8/100

GPT-6 Astra: Raccoon SVG benchmark
Researcher observations

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.

Benchmark task

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.

MISSINGNO.57.5/100

MISSINGNO.: Raccoon SVG benchmark
Researcher observations

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.

Benchmark task

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.

Compare category scores
CategoryGPT-6 AstraMISSINGNO.
Task & likeness8.5/108.0/10
Anatomy & construction7.5/105.0/10
Space & placement8.0/105.5/10
Light & reflections6.5/105.0/10
Purposeful detail8.0/106.0/10
Artistry & character8.5/105.0/10
Materials & effects8.0/104.5/10
Visual integrity8.5/107.0/10

Pelican

Different task instructions

GPT-6 Astra75.0/100

GPT-6 Astra: Pelican SVG benchmark
Researcher observations

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.

Benchmark task

A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.

MISSINGNO.58.8/100

MISSINGNO.: Pelican SVG benchmark
Researcher observations

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.

Benchmark task

A cycling pelican tests anatomy, riding posture and contact between the bird and the bicycle.

Compare category scores
CategoryGPT-6 AstraMISSINGNO.
Task & likeness8.0/107.0/10
Anatomy & construction6.5/105.5/10
Space & placement7.5/106.0/10
Light & reflections7.0/105.0/10
Purposeful detail8.0/105.5/10
Artistry & character8.0/106.0/10
Materials & effects7.5/104.5/10
Visual integrity8.0/106.5/10