local/sefi-image-5b-turbo Benchmark — 42/100 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.

MakerSefi

FamilySefi Image

Model size5B5B params

Costself-hostedfree

Latency5.9s median

Run targetlocal/sefi-image-5b-turbo

Effective requestmodel: sefi-image-5b-turbo · 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

49%

Typography Style67%

Writing accuracy33%

Spatial Reasoning

50%

Attributes Binding56%

Compositionality44%

Counting56%

Negation56%

Relative Position50%

Scale & Proportions22%

Human realism

57%

Faces & Expressions50%

Full Body42%

Hands42%

Multi-Subject50%

Professional Studio

48%

Camera & Lighting42%

Color Precision50%

Photorealism33%

Graphical design

50%

Data Visualisation33%

Layout & Design44%

Style Diversity58%

Truthfulness

61%

Photorealism67%

Physics & Reflections58%

World Knowledge42%