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-max

API

FLUX.2 Max is Black Forest Labs' higher-capacity FLUX.2 tier, described as a top-end generation model with support for stronger visual output and, when prompted, grounding search for current visual context.

Maker
Black Forest Labs
Family
FLUX.2
Model Size
not disclosed
unknown
Cost
from $0.07 / image
high
Run Target
bfl/flux-2-max
Effective Request
aspect_ratio: 1:1
59.6
Overall
63%
Capability
56.6
Est. Preference
120
Pass
72
Fail
26.7s
Avg Latency
14.3s
Min Latency
60.4s
Max Latency
Text Rendering93%Spatial Reasoning75%Human realism24%Truthfulness41%Professional Studio89%Graphical design75%Preference57%Latency3%

All 192 generations

Text Rendering93%

Typography Style100%

Writing accuracy92%

Spatial Reasoning75%

Attributes Binding89%

Compositionality89%

Counting44%

Negation89%

Relative Position83%

Scale & Proportions56%

Human realism24%

Faces & Expressions83%

Full Body0%

Hands0%

Multi-Subject0%

Truthfulness41%

Photorealism0%

Physics & Reflections50%

World Knowledge42%

Professional Studio89%

Camera & Lighting92%

Color Precision83%

Photorealism100%

Graphical design75%

Data Visualisation0%

Layout & Design78%

Style Diversity92%