STEALTHMARK · HEAD TO HEAD

All comparisons ↗

Kimi K3 vs Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) leads Kimi K3 by 12.0 points on StealthMark, 62.6 to 50.6. It leads on 5 of 9 benchmarks, including Foresight, and Kimi K3 on 4, including Latch. Union Alpha scores higher on Newton's Mirror, with its largest category advantage in reflections.

See the difference ↘
All results ↗
Model guide ↗

Overall StealthMark

50.6/100Overall rank 19/58
62.6/100Overall rank 8/58

General Intelligence

44.4/100
58.7/100

Visual Intelligence

61.4/100
69.6/100

Reasoning efficiency

How many reasoning tokens a model needs for the quality it delivers.

Single-step reasoning How much it reasons to understand a task

Multi-step reasoning How much it keeps reasoning as a task goes on

  • Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)Not measured

Cost efficiency

How much Performance you get for your money.

Open or local

Open modelKimi K3 License

BENCHMARK BY BENCHMARK

Where each model leads

  • Kimi K3
  • Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
  • Other models
  • Median

DATASHEET

Specs and pricing

Kimi K3 takes 4× more context and runs 4.5× faster in our speed test. Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) scores higher on StealthMark.

  • Kimi K3
  • Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
  • Other models
  • Median
PricePer 1M output tokens, input below
Kimi K3: $10.80$0.24 inputPareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM): No list price
Context windowTokens per request, max output below
Kimi K3: 1M4× larger1M max outputPareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM): 262K131K max output
Output speedTokens per second, our measurement
Kimi K3: 1154.5× fastertokens/sPareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM): 26tokens/s
SizeTotal and active parameters
Kimi K3: 2.8T104B activePareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM): Not disclosed
Developer
Moonshot AICircuit & Chisel / Unbiased
Released
16 Jul 202620 Sep 2026
Origin
ChinaUnited States
Input → output
Text, image → textBoth
Architecture
Mixture-of-Experts with KDA + Gated MLA attention, 896 experts and native vision encoderNot published

Full datasheets, every price and sourceKimi K3 ↗Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) ↗

Anticipating threats and planning ahead. Collect coins while avoiding ghosts. Ghosts move before the player each turn, and collisions cost a life. 3 levels in sequence, up to 66 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.

Kimi K3 52.7/100Rank 19/54

Survived 0 of 3
Foresight: complete playing field
Level 311 ptsTurn 17 / 20

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) 78.7/100Rank 1/54

Survived 2 of 3
Foresight: complete playing field
Level 319 ptsTurn 20 / 20

Points per level

LevelKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
Level 1Second Thoughts
17 ptsOut of lives
27 ptsSurvived
Level 2Pressure Pot
10 ptsOut of lives
12 ptsOut of lives
Level 3Hide or Flight
11 ptsOut of lives
19 ptsSurvived

Cooperative planning and learning from experience. Coordinate two robots to push numbered cores onto matching sockets. Pressure plates hold doors open; moving either robot uses one turn. 5 levels in sequence, up to 152 turns in total. Earlier moves and outcomes carry forward, testing learning across levels.

Kimi K3 57.7/100Rank 27/55

Cleared 3 of 5
Latch: complete playing field
Level 514.2 ptsTurn 4 / 40

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) 36.5/100Rank 37/55

Cleared 4 of 5
Latch: complete playing field
Level 519.4 ptsTurn 33 / 40

Points per level

LevelKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
Level 1Hold the door
5.2 ptsCleared
5.2 ptsCleared
Level 2Crossfeed
9.6 ptsCleared
9.8 ptsCleared
Level 3Relay vault
18.2 ptsClearedFinished early · 3 beacons left
14.2 ptsClearedFinished early · 4 beacons left
Level 4Passing bay
14.2 ptsFinished early
9.2 ptsFinished early
Level 5Second thoughts
14.2 ptsFinished early
19.4 ptsCleared

Newton's Mirror

Same task instructions

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

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) 52.0/100 Rank 8/58

Union Alpha: Newton's Mirror SVG benchmark

The judgment

Kimi K3

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.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

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.

Score breakdown

CategoryKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
State at 600 ms18.0/2524.0/25
Camera and placement8.0/1513.5/15
Reflections24.0/12034.5/120
Materials and scale6.5/1010.0/10
Floor-damage reward0.0/100.0/10

Benchmark task

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

Selfie

Different task instructions

Visual construction and creative interpretation.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) 76.5/100 Rank 6/58

Union Alpha: Selfie SVG benchmark

The judgment

Kimi K3

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.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

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.

Score breakdown

CategoryKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
Task & likeness9.0/109.0/10
Anatomy & construction7.0/107.5/10
Space & placement8.0/108.5/10
Light & reflections7.5/107.5/10
Purposeful detail8.0/107.0/10
Artistry & character8.5/107.5/10
Materials & effects7.5/107.5/10
Visual integrity8.0/107.0/10

Benchmark task

Kimi K3

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.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

An interwoven metallic emblem tests ribbon geometry, overlapping forms and the appearance of reflective metal against a dark background.

Mona Lisa

Different task instructions

Precise visual recall and portrait reconstruction.

The judgment

Kimi K3

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.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

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.

Score breakdown

CategoryKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
Task & likeness4.5/106.5/10
Anatomy & construction3.5/105.0/10
Space & placement4.5/106.5/10
Light & reflections4.5/106.0/10
Purposeful detail4.0/105.5/10
Artistry & character4.5/106.5/10
Materials & effects4.5/105.0/10
Visual integrity5.0/107.0/10

Benchmark task

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

Beach

Different task instructions

Creativity, perspective and visual beauty.

The judgment

Kimi K3

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.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

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.

Score breakdown

CategoryKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
Task & likeness8.5/109.0/10
Anatomy & construction7.0/107.0/10
Space & placement7.0/108.0/10
Light & reflections6.5/107.0/10
Purposeful detail8.5/107.0/10
Artistry & character8.5/106.0/10
Materials & effects6.5/106.5/10
Visual integrity7.0/107.0/10

Benchmark task

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

Ground Truth

Same task instructions

Physical reasoning, light and visual reconstruction.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM) 63.3/100 Rank 9/58

Union Alpha: Ground Truth SVG benchmark

The judgment

Kimi K3

This benchmark has not been assessed yet.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

This benchmark has not been assessed yet.

Benchmark task

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

Raccoon

Different task instructions

Scene composition and interacting subjects.

The judgment

Kimi K3

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.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

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.

Score breakdown

CategoryKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
Task & likeness8.0/108.0/10
Anatomy & construction6.5/106.5/10
Space & placement7.5/107.5/10
Light & reflections7.0/105.5/10
Purposeful detail7.5/107.0/10
Artistry & character6.0/107.0/10
Materials & effects7.5/107.5/10
Visual integrity7.0/107.0/10

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.

Pelican

Different task instructions

Subject construction and spatial relationships.

The judgment

Kimi K3

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.

Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)

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.

Score breakdown

CategoryKimi K3Pareto 26.9 (Union of GPT-6 Astra + DeepSeek + GLM)
Task & likeness8.0/108.0/10
Anatomy & construction6.5/106.0/10
Space & placement7.5/107.0/10
Light & reflections5.0/106.0/10
Purposeful detail6.5/106.0/10
Artistry & character6.0/106.0/10
Materials & effects6.0/106.0/10
Visual integrity6.5/108.5/10

Benchmark task

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