STEALTHMARK · HEAD TO HEAD

All comparisons ↗

GLM-5.3-Flash vs Step 5 Preview

GLM-5.3-Flash leads Step 5 Preview by 9.5 points on StealthMark, 48.5 to 39.0. It leads on 7 of 9 benchmarks, including Foresight, and Step 5 Preview on 2, including Latch. Ox Alpha scores higher on Ground Truth.

See the difference ↘
Model guide ↗
All results ↗

Overall StealthMark

48.5/100Overall rank 19/53
39.0/100Overall rank 36/53

General Intelligence

42.3/100
37.1/100

Visual Intelligence

59.4/100
41.9/100

Reasoning efficiency

On every task, result quality and reasoning tokens are each compared with all measured models and weighed equally. Each dot is one model.

Single-step reasoning Result quality weighed against reasoning tokens on a single problem

Multi-step reasoning Result quality weighed against reasoning tokens across connected steps, as in agentic workflows

Open or local

Local modelOpen model · MIT

BENCHMARK BY BENCHMARK

Where each model leads

  • GLM-5.3-Flash
  • Step 5 Preview
  • Other models
  • Median

DATASHEET

Specs and pricing

Both take about 1M tokens of context.

  • GLM-5.3-Flash
  • Step 5 Preview
  • Other models
  • Median
PricePer 1M output tokens, input below
GLM-5.3-Flash: $0.50$0.15 inputStep 5 Preview: No list price
Context windowTokens per request, max output below
GLM-5.3-Flash: 1MTied131K max outputStep 5 Preview: 1MTied
Output speedTokens per second, our measurement
GLM-5.3-Flash: 89tokens/sStep 5 Preview: Not measured
SizeTotal and active parameters
GLM-5.3-Flash: 320B18B activeStep 5 Preview: 600B27B active
Developer
Z.AIStepFun
Released
26 Aug 2026Not published
Input → output
Text, image → textText, image, video → text
Architecture
Hybrid sparse and linear attention MoESparse mixture-of-experts; 92 layers (88 MoE), 352 routed experts with 8 selected per token plus a shared expert; grouped-query attention with 512-token sliding windows and full attention every fourth layer; 3 multi-token-prediction layers; vision tower for image and video input

Full datasheets, every price and sourceGLM-5.3-Flash ↗Step 5 Preview ↗

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-Flash 47.0/100Rank 20/49

Survived 1 of 3
Foresight: complete playing field
Level 316 ptsTurn 20 / 20

Step 5 Preview 59.9/100Rank 42/491 of 3 levels

1 of 3 played
Foresight: complete playing field
Level 122 ptsTurn 28 / 28

Points per level

LevelGLM-5.3-FlashStep 5 Preview
Level 1Second Thoughts
9 ptsOut of lives
22 ptsSurvived
Level 2Pressure Pot
6 ptsOut of lives
Pending
Level 3Hide or Flight
16 ptsSurvived
Pending

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-Flash 4.0/100Rank 50/503 of 5 levels

3 of 5 played
Latch: complete playing field
Level 30.6 ptsTurn 33 / 36

Step 5 Preview 40.7/100Rank 28/50

Cleared 4 of 5
Latch: complete playing field
Level 515 ptsTurn 35 / 40

Points per level

LevelGLM-5.3-FlashStep 5 Preview
Level 1Hold the door
4.4 ptsCleared
5.2 ptsCleared
Level 2Crossfeed
5.2 ptsFinished early
9 ptsCleared
Level 3Relay vault
0.6 ptsFinished early
18.6 ptsCleared
Level 4Passing bay
Pending
15.6 ptsClearedFinished early · 3 beacons left
Level 5Second thoughts
Not played
15 ptsFinished early

Newton's Mirror

Same task instructions

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

The judgment

GLM-5.3-Flash

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.

Step 5 Preview

The main sphere, airborne black and small colored features occupy broadly appropriate positions. Most of the defining mirror system is absent, and paint interaction remains only partially supported. The small paint features do not visibly establish a separate volume-scale violation.

Score breakdown

CategoryGLM-5.3-FlashStep 5 Preview
State at 600 ms14.0/2520.0/25
Camera and placement9.0/1513.5/15
Reflections21.5/12020.0/120
Materials and scale3.5/107.0/10
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-Flash

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.

Step 5 Preview

A charming diagonal climber and warm palette communicate the central idea. Simple rounded construction, crowded limb counting, uncertain footholds and an unattached upper shelf limit its physical coherence. Growth bands and the yarn trail add specificity, but the requested material distinctions and unravelled bootie remain underdeveloped.

Score breakdown

CategoryGLM-5.3-FlashStep 5 Preview
Task & likeness8.5/106.0/10
Anatomy & construction7.5/104.5/10
Space & placement7.5/105.0/10
Light & reflections8.0/105.5/10
Purposeful detail7.5/105.5/10
Artistry & character7.5/105.0/10
Materials & effects7.5/106.0/10
Visual integrity7.0/105.0/10

Benchmark task

GLM-5.3-Flash

A rounded jade bull raising a hoof tests anatomy, balance and the appearance of polished stone and metal.

Step 5 Preview

Character scenes test expression, anatomy, materials and coherent visual construction.

Mona Lisa

Different task instructions

Precise visual recall and portrait reconstruction.

The judgment

GLM-5.3-Flash

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.

Step 5 Preview

A recognizable, coherent simplification with purposeful facial contours and differentiated fingers. These provide more construction and detail than the broad silhouette alone suggests. The oversized face, frontal seated pose, unresolved wrists and sparse landscape still substantially limit likeness, depth and finish.

Score breakdown

CategoryGLM-5.3-FlashStep 5 Preview
Task & likeness4.5/104.0/10
Anatomy & construction4.5/104.0/10
Space & placement6.0/104.0/10
Light & reflections6.0/103.5/10
Purposeful detail5.0/104.0/10
Artistry & character5.0/104.0/10
Materials & effects5.5/103.0/10
Visual integrity4.0/103.5/10

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

GLM-5.3-Flash

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.

Step 5 Preview

A recognizable sunset beach with clear warm-cool contrast and playful accessories. Basic animal construction, shallow foreground placement and schematic reflections limit spatial and physical credibility. Sparse descriptive detail and abruptly ended sky bands constrain the finish.

Score breakdown

CategoryGLM-5.3-FlashStep 5 Preview
Task & likeness8.0/104.0/10
Anatomy & construction7.0/103.0/10
Space & placement8.0/103.5/10
Light & reflections8.0/103.5/10
Purposeful detail8.0/103.0/10
Artistry & character8.5/103.0/10
Materials & effects6.5/103.5/10
Visual integrity8.0/104.0/10

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.

The judgment

GLM-5.3-Flash

This benchmark has not been assessed yet.

Step 5 Preview

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

GLM-5.3-Flash

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.

Step 5 Preview

Recognizable animals, a bright palette and identifying accessories support a basic cartoon. Primitive anatomy, partial seated support, unresolved steering and weak hull-water contact limit the central action. Lighting and spray provide atmosphere without establishing convincing physical depth.

Score breakdown

CategoryGLM-5.3-FlashStep 5 Preview
Task & likeness7.0/104.0/10
Anatomy & construction5.0/104.5/10
Space & placement6.0/103.0/10
Light & reflections5.0/103.5/10
Purposeful detail6.5/104.0/10
Artistry & character4.0/103.5/10
Materials & effects6.0/103.0/10
Visual integrity6.0/103.0/10

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

GLM-5.3-Flash

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.

Step 5 Preview

The pelican and bicycle are recognizable, with a pleasant palette and basic diffuse shading. The riding relationship fails physically: the bird is unsupported, the front wheel is unmounted, and pedalling is unresolved. Simple geometric construction and sparse material treatment further limit the drawing.

Score breakdown

CategoryGLM-5.3-FlashStep 5 Preview
Task & likeness8.5/103.0/10
Anatomy & construction7.5/103.0/10
Space & placement7.5/103.0/10
Light & reflections6.5/103.0/10
Purposeful detail7.0/104.0/10
Artistry & character8.0/103.0/10
Materials & effects6.5/103.0/10
Visual integrity8.0/102.0/10

Benchmark task

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