local/z-image-6b Benchmark — 47/100 Overall, 192 Prompts, Every Image Shown

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 run.

MakerTongyi

FamilyZ-Image

Model size6B6B params

Costself-hostedfree

Latency131s median

Run targetlocal/z-image-6b

Effective requestmodel: z-image-6b · size: 1024x1024 · seed: 42

SourcesModel cardWeights & checkpoint

All 192 generations

Every challenge, grouped by category and subcategory. Solid-bordered tiles with a check passed the VLM judge; dashed tiles with a cross failed. Images are placeholders until generated.

Text Rendering

62%

Typography Style33%

Writing accuracy58%

Spatial Reasoning

47%

Attributes Binding44%

Compositionality33%

Counting44%

Negation11%

Relative Position50%

Scale & Proportions67%

Human realism

68%

Faces & Expressions67%

Full Body67%

Hands58%

Multi-Subject67%

Professional Studio

53%

Camera & Lighting33%

Color Precision58%

Photorealism67%

Graphical design

47%

Data Visualisation67%

Layout & Design33%

Style Diversity42%

Truthfulness

42%

Photorealism33%

Physics & Reflections42%

World Knowledge50%