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.

openai/gpt-image-2.5-flare

API

GPT Image 2.5 Flare is one of two API variants OpenAI released alongside ChatGPT Images 2.5 (September 2026), positioned as OpenAI's fastest model for high-quality, everyday image generation.

Maker
OpenAI
Family
GPT Image
Model Size
not disclosed
unknown
Cost
token-based
high
Run Target
openai/gpt-image-2.5-flare
Effective Request
model: gpt-image-2.5-flare · size: 1024x1024 · quality: auto
81.5
Overall
87%
Capability
76.5
Est. Preference
166
Pass
26
Fail
13.2s
Avg Latency
7.7s
Min Latency
29.8s
Max Latency
Text Rendering100%Spatial Reasoning95%Human realism55%Truthfulness85%Professional Studio100%Graphical design100%Preference77%Latency24%

All 192 generations

Text Rendering100%

Typography Style100%

Writing accuracy100%

Spatial Reasoning95%

Attributes Binding100%

Compositionality100%

Counting100%

Negation100%

Relative Position100%

Scale & Proportions67%

Human realism55%

Faces & Expressions92%

Full Body8%

Hands67%

Multi-Subject50%

Truthfulness85%

Photorealism100%

Physics & Reflections75%

World Knowledge92%

Professional Studio100%

Camera & Lighting100%

Color Precision100%

Photorealism100%

Graphical design100%

Data Visualisation100%

Layout & Design100%

Style Diversity100%