local/prxpixel-t2i-7b Benchmark — 46/100 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.

MakerPrxPixel

FamilyPrxPixel

Model size7B7B params

Costself-hostedfree

Latency64.7s median

Run targetlocal/prxpixel-t2i-7b

Effective requestmodel: prxpixel-t2i-7b · 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

45%

Typography Style33%

Writing accuracy33%

Spatial Reasoning

55%

Attributes Binding67%

Compositionality22%

Counting78%

Negation56%

Relative Position42%

Scale & Proportions44%

Human realism

29%

Faces & Expressions17%

Full Body17%

Hands17%

Multi-Subject33%

Professional Studio

50%

Camera & Lighting25%

Color Precision42%

Photorealism33%

Graphical design

45%

Data Visualisation33%

Layout & Design33%

Style Diversity25%

Truthfulness

33%

Photorealism33%

Physics & Reflections25%

World Knowledge25%