STEALTHMARK · HEAD TO HEAD

All comparisons ↗

GLM-5.3 vs GPT-6 Astra

GPT-6 Astra leads GLM-5.3 by 21.7 points on StealthMark, 74.6 to 52.9. It leads on all 9 benchmarks. 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

52.9/100Overall rank 16/58
74.6/100Overall rank 2/58

General Intelligence

46.5/100
73.4/100

Visual Intelligence

64.1/100
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

  • GPT-6 Astra

Cost efficiency

How much Performance you get for your money.

Open or local

Open modelGLM-5.3 License

BENCHMARK BY BENCHMARK

Where each model leads

  • GLM-5.3
  • GPT-6 Astra
  • Other models
  • Median

DATASHEET

Specs and pricing

GLM-5.3 is 11× cheaper per output token and runs 32% faster in our speed test. GPT-6 Astra scores higher on StealthMark. Both take about 1M tokens of context.

  • GLM-5.3
  • GPT-6 Astra
  • Other models
  • Median
PricePer 1M output tokens, input below
GLM-5.3: $4.4011× cheaper$1.40 inputGPT-6 Astra: $50$10 input
Context windowTokens per request, max output below
GLM-5.3: 1MTied128K max outputGPT-6 Astra: 1.05MTied128K max output
Output speedTokens per second, our measurement
GLM-5.3: 8632% fastertokens/sGPT-6 Astra: 65tokens/s
SizeTotal and active parameters
GLM-5.3: 753BAll activeGPT-6 Astra: Not disclosed
Developer
Z.AIOpenAI
Released
17 Aug 20263 Sep 2026
Origin
ChinaUnited States
Input → output
Text → textText, image → text
Architecture
GLM MoE with DSA sparse attentionNot published

Full datasheets, every price and sourceGLM-5.3 ↗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.

GLM-5.3 43.6/100Rank 25/54

Survived 1 of 3
Foresight: complete playing field
Level 34 ptsTurn 7 / 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

LevelGLM-5.3GPT-6 Astra
Level 1Second Thoughts
27 ptsSurvived
27 ptsSurvived
Level 2Pressure Pot
6 ptsOut of lives
11 ptsSurvived
Level 3Hide or Flight
4 ptsOut of lives
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.

GLM-5.3 25.6/100Rank 44/55

Cleared 2 of 5
Latch: complete playing field
Level 512.2 ptsTurn 14 / 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

LevelGLM-5.3GPT-6 Astra
Level 1Hold the door
5.2 ptsCleared
5.2 ptsCleared
Level 2Crossfeed
9 ptsCleared
10.2 ptsClearedFinished early · 1 beacon left
Level 3Relay vault
7.6 ptsFinished early
19.2 ptsCleared
Level 4Passing bay
3.2 ptsFinished early
18.2 ptsClearedFinished early · 3 beacons left
Level 5Second thoughts
12.2 ptsFinished early
20.6 ptsCleared

Newton's Mirror

Same task instructions

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

The judgment

GLM-5.3

The central sphere, isolated airborne black blob and crimson leading-end placement are well established. Recent cyan contact remains only partially demonstrated. The dominant weakness is optical: sparse shaded sphere rows, missing outer chains and diffuse highlights replace extensive mirror repetition and curved nested imagery.

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

CategoryGLM-5.3GPT-6 Astra
State at 600 ms22.5/2525.0/25
Camera and placement13.0/1515.0/15
Reflections37.0/12070.0/120
Materials and scale7.0/1010.0/10
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

GLM-5.3

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.

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

CategoryGLM-5.3GPT-6 Astra
Task & likeness9.0/109.0/10
Anatomy & construction7.0/108.0/10
Space & placement7.0/107.5/10
Light & reflections8.0/107.0/10
Purposeful detail8.0/108.0/10
Artistry & character7.0/108.5/10
Materials & effects7.5/108.0/10
Visual integrity8.0/108.5/10

Benchmark task

GLM-5.3

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.

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

Same task instructions

Precise visual recall and portrait reconstruction.

The judgment

GLM-5.3

A recognizable, coherent interpretation with purposeful facial accents, readable shading and intact principal attachments. These demonstrated strengths warrant modest upward calibration. Frontal seating, distorted proportions, schematic hands and uniform surfaces still limit likeness, anatomy, depth and artistic refinement.

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

CategoryGLM-5.3GPT-6 Astra
Task & likeness4.5/106.5/10
Anatomy & construction4.5/106.5/10
Space & placement4.5/106.5/10
Light & reflections6.0/106.5/10
Purposeful detail5.0/106.5/10
Artistry & character5.0/107.0/10
Materials & effects5.0/107.0/10
Visual integrity6.0/107.5/10

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

GLM-5.3

A coherent beach illustration with a clear sunset, playful subjects and purposeful small details. Varied silhouettes and broken reflections strengthen its composition. Basic human anatomy, shallow foreground construction and limited material differentiation constrain the result. The kite line reaches the smaller figure rather than requiring a grip from the nearby animal.

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

CategoryGLM-5.3GPT-6 Astra
Task & likeness8.5/109.0/10
Anatomy & construction5.5/107.5/10
Space & placement6.0/108.5/10
Light & reflections7.0/107.5/10
Purposeful detail7.0/108.5/10
Artistry & character5.5/108.5/10
Materials & effects6.5/107.5/10
Visual integrity7.0/107.5/10

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

GLM-5.3

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

GLM-5.3

A coherent sunset illustration with readable steering, expressive accessories and effective translucent water effects. Its strongest physical lighting evidence is the aligned sun reflection. The raccoon's seated support remains unresolved, while simplified joints and limited viewpoint integration constrain construction and artistry.

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

CategoryGLM-5.3GPT-6 Astra
Task & likeness7.0/108.5/10
Anatomy & construction6.0/107.5/10
Space & placement6.0/107.5/10
Light & reflections7.0/106.5/10
Purposeful detail7.5/108.0/10
Artistry & character6.0/107.5/10
Materials & effects7.5/107.5/10
Visual integrity5.5/108.0/10

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

GLM-5.3

A coherent coastal illustration with a distinctive pelican silhouette, expressive flowing shapes and substantial bicycle detail. Feet meet plausible pedal platforms, but steering contact remains unresolved. Simplified leg articulation, modest depth and approximate directional lighting limit physical construction. The apparent neck opening is ambiguous rather than a verified break.

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

CategoryGLM-5.3GPT-6 Astra
Task & likeness7.0/108.0/10
Anatomy & construction5.0/107.0/10
Space & placement6.0/107.5/10
Light & reflections6.0/107.0/10
Purposeful detail7.5/108.0/10
Artistry & character6.0/108.0/10
Materials & effects6.5/107.5/10
Visual integrity6.0/108.0/10

Benchmark task

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