ImageBench

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.

bfl/flux-2-pro

API

FLUX.2 Pro is a hosted Black Forest Labs FLUX.2 model for premium image generation workloads, using the same FLUX.2 family emphasis on prompt following, image quality, and production-oriented generation.

Maker
Black Forest Labs
Family
FLUX.2
Model Size
not disclosed
unknown
Cost
from $0.03 / image
high
Run Target
bfl/flux-2-pro
Effective Request
aspect_ratio: 1:1
60.1
Overall
63%
Capability
57.3
Est. Preference
121
Pass
71
Fail
11.8s
Avg Latency
7.3s
Min Latency
67.1s
Max Latency
Text Rendering80%Spatial Reasoning75%Human realism31%Truthfulness44%Professional Studio85%Graphical design75%Preference57%Latency27%

All 192 generations

Text Rendering80%

Typography Style100%

Writing accuracy75%

Spatial Reasoning75%

Attributes Binding78%

Compositionality89%

Counting56%

Negation89%

Relative Position67%

Scale & Proportions78%

Human realism31%

Faces & Expressions67%

Full Body25%

Hands8%

Multi-Subject17%

Truthfulness44%

Photorealism33%

Physics & Reflections50%

World Knowledge42%

Professional Studio85%

Camera & Lighting83%

Color Precision92%

Photorealism67%

Graphical design75%

Data Visualisation0%

Layout & Design78%

Style Diversity92%