STEALTHMARK · HEAD TO HEAD

All comparisons ↗

Claude Sonnet 5.5 vs GPT-6 Astra

GPT-6 Astra leads Claude Sonnet 5.5 by 4.4 points on StealthMark, 74.6 to 70.2. It leads on 6 of 9 benchmarks, including both games, and Claude Sonnet 5.5 on 3. GPT-6 Astra scores higher on Newton's Mirror, with its largest category advantage in reflections.

See the difference ↘
All results ↗
All results ↗

Overall StealthMark

70.2/100Overall rank 4/58
74.6/100Overall rank 2/58

General Intelligence

67.0/100
73.4/100

Visual Intelligence

76.1/100
76.9/100

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

  • Claude Sonnet 5.5
  • GPT-6 Astra

Cost efficiency

How much Performance you get for your money.

BENCHMARK BY BENCHMARK

Where each model leads

  • Claude Sonnet 5.5
  • GPT-6 Astra
  • Other models
  • Median

DATASHEET

Specs and pricing

Claude Sonnet 5.5 is 5× cheaper per output token and runs 3.8× faster in our speed test. GPT-6 Astra scores higher on StealthMark. Both take about 1M tokens of context.

  • Claude Sonnet 5.5
  • GPT-6 Astra
  • Other models
  • Median
PricePer 1M output tokens, input below
Claude Sonnet 5.5: $105× cheaper$2 inputGPT-6 Astra: $50$10 input
Context windowTokens per request, max output below
Claude Sonnet 5.5: 1MTied128K max outputGPT-6 Astra: 1.05MTied128K max output
Output speedTokens per second, our measurement
Claude Sonnet 5.5: 2493.8× fastertokens/sGPT-6 Astra: 65tokens/s
Developer
AnthropicOpenAI
Released
28 Sep 20263 Sep 2026
Origin
United StatesBoth
Input → output
Text, image → textBoth

Full datasheets, every price and sourceClaude Sonnet 5.5 ↗GPT-6 Astra ↗

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.

Claude Sonnet 5.5 59.6/100Rank 13/54

Survived 2 of 3
Foresight: complete playing field
Level 310 ptsTurn 20 / 20

GPT-6 Astra 76.1/100Rank 3/54

Survived 3 of 3
Foresight: complete playing field
Level 318 ptsTurn 20 / 20

Points per level

LevelClaude Sonnet 5.5GPT-6 Astra
Level 1Second Thoughts
25 ptsSurvived
27 ptsSurvived
Level 2Pressure Pot
10 ptsOut of lives
11 ptsSurvived
Level 3Hide or Flight
10 ptsSurvived
18 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.

Claude Sonnet 5.5 83.9/100Rank 5/55

Cleared 5 of 5
Latch: complete playing field
Level 520.6 ptsTurn 27 / 40

GPT-6 Astra 87.6/100Rank 2/55

Cleared 5 of 5
Latch: complete playing field
Level 520.6 ptsTurn 27 / 40

Points per level

LevelClaude Sonnet 5.5GPT-6 Astra
Level 1Hold the door
5.2 ptsCleared
5.2 ptsCleared
Level 2Crossfeed
10 ptsClearedFinished early · 1 beacon left
10.2 ptsClearedFinished early · 1 beacon left
Level 3Relay vault
19.2 ptsCleared
19.2 ptsCleared
Level 4Passing bay
18.2 ptsClearedFinished early · 3 beacons left
18.2 ptsClearedFinished early · 3 beacons left
Level 5Second thoughts
20.6 ptsCleared
20.6 ptsCleared

Newton's Mirror

Same task instructions

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

The judgment

Claude Sonnet 5.5

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.

GPT-6 Astra

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.

Score breakdown

CategoryClaude Sonnet 5.5GPT-6 Astra
State at 600 ms24.0/2525.0/25
Camera and placement15.0/1515.0/15
Reflections35.0/12070.0/120
Materials and scale10.0/1010.0/10
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.

The judgment

Claude Sonnet 5.5

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.

GPT-6 Astra

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.

Score breakdown

CategoryClaude Sonnet 5.5GPT-6 Astra
Task & likeness8.0/109.0/10
Anatomy & construction8.0/108.0/10
Space & placement8.0/107.5/10
Light & reflections7.5/107.0/10
Purposeful detail8.0/108.0/10
Artistry & character8.0/108.5/10
Materials & effects8.0/108.0/10
Visual integrity8.5/108.5/10

Benchmark task

Claude Sonnet 5.5

A coral squid with spreading arms tests connected anatomy, balanced composition and soft surface shading.

GPT-6 Astra

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.

Mona Lisa

Different task instructions

Precise visual recall and portrait reconstruction.

The judgment

Claude Sonnet 5.5

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.

GPT-6 Astra

Controlled hair contours, neckline detail, modeled sleeves and a layered river landscape give this recognizable portrait developed illustrative finish. The elongated face, largely frontal torso, crowded fingers and weakly established chair support limit replication, anatomical precision and spatial conviction. Small construction slips at the neck and jaw, a stray dark shape under the right cuff and an abruptly ending armrest bar weaken integrity.

Score breakdown

CategoryClaude Sonnet 5.5GPT-6 Astra
Task & likeness6.5/106.5/10
Anatomy & construction5.0/106.5/10
Space & placement6.0/106.5/10
Light & reflections6.5/106.5/10
Purposeful detail6.0/106.5/10
Artistry & character6.5/107.0/10
Materials & effects6.0/107.0/10
Visual integrity6.5/107.5/10

Benchmark task

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

Beach

Same task instructions

Creativity, perspective and visual beauty.

The judgment

Claude Sonnet 5.5

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.

GPT-6 Astra

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.

Score breakdown

CategoryClaude Sonnet 5.5GPT-6 Astra
Task & likeness9.0/109.0/10
Anatomy & construction8.0/107.5/10
Space & placement8.5/108.5/10
Light & reflections8.5/107.5/10
Purposeful detail9.5/108.5/10
Artistry & character9.5/108.5/10
Materials & effects8.5/107.5/10
Visual integrity8.0/107.5/10

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.

The judgment

Claude Sonnet 5.5

This benchmark has not been assessed yet.

GPT-6 Astra

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

Same task instructions

Scene composition and interacting subjects.

The judgment

Claude Sonnet 5.5

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.

GPT-6 Astra

A controlled, expressive illustration with readable riding and steering, articulated poses and purposeful small details. Layered foam and coherent hull perspective establish motion and depth. Simplified paws, compressed control fittings and limited directional lighting constrain its physical sophistication.

Score breakdown

CategoryClaude Sonnet 5.5GPT-6 Astra
Task & likeness8.5/108.5/10
Anatomy & construction7.0/107.5/10
Space & placement8.0/107.5/10
Light & reflections6.5/106.5/10
Purposeful detail8.0/108.0/10
Artistry & character7.0/107.5/10
Materials & effects7.5/107.5/10
Visual integrity8.0/108.0/10

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

Same task instructions

Subject construction and spatial relationships.

The judgment

Claude Sonnet 5.5

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.

GPT-6 Astra

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.

Score breakdown

CategoryClaude Sonnet 5.5GPT-6 Astra
Task & likeness9.0/108.0/10
Anatomy & construction7.5/107.0/10
Space & placement7.5/107.5/10
Light & reflections6.5/107.0/10
Purposeful detail8.0/108.0/10
Artistry & character7.5/108.0/10
Materials & effects6.5/107.5/10
Visual integrity8.0/108.0/10

Benchmark task

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