ImageBench V1 —

192 evaluations across 6 categories

Benchmark V1 verdicts are produced by VLM judges and can contain mistakes. Treat PASS/FAIL labels as machine-assisted assessments, and inspect the images yourself. Learn more about the methodology.

Generation Details

Source-backed model context, size, cost, and request settings for this ImageBench V1 run.

xai/grok-imagine-2

API

Grok Imagine Image 2.0 is xAI's text-to-image model (released Aug 2026), emphasizing photorealism, natural lighting, physics, and textures. Served via xAI's first-party API (OpenAI-images-compatible). The first xAI model on this board.

Maker: xAI
Family: Grok Imagine 2.0
Model Size: not disclosed
unknown
Cost: $0.04 / image
high
Run Target: xai/grok-imagine-image-2.0
Effective Request: aspect_ratio: 1:1 · resolution: 1k
Sources: Grok Imagine Image 2.0 — xAI xAI image generation API docs

73.6

Overall

76%

Capability

71.6

Est. Preference

145

Pass

47

Fail

69.0s

Avg Latency

24.0s

Min Latency

115.0s

Max Latency

All 192 generations

Text Rendering100%

Typography Style100%

Writing accuracy100%

Spatial Reasoning89%

Attributes Binding100%

Compositionality100%

Counting100%

Negation100%

Relative Position75%

Scale & Proportions67%

Human realism38%

Faces & Expressions83%

Full Body8%

Hands17%

Multi-Subject50%

Truthfulness56%

Photorealism100%

Physics & Reflections42%

World Knowledge58%

Professional Studio100%

Camera & Lighting100%

Color Precision100%

Photorealism100%

Graphical design88%

Data Visualisation67%

Layout & Design100%

Style Diversity83%