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.

fal/bytedance/seedream-v4

API

Seedream 4.0 is ByteDance Seed's unified image creation model for text-to-image and editing, aimed at multimodal generation, reference consistency, knowledge-based prompts, and high-resolution output.

Maker: ByteDance Seed
Family: Seedream
Model Size: not disclosed
unknown
Cost: $0.03 / image
high
Run Target: fal/fal-ai/bytedance/seedream/v4/text-to-image
Effective Request: image_size: square_hd · num_images: 1 · seed: 7
Sources: ByteDance Seedream 4.0 Seedream 4.0 on fal fal pricing

59.4

Overall

63%

Capability

56.2

Est. Preference

120

Pass

72

Fail

14.1s

Avg Latency

7.7s

Min Latency

38.0s

Max Latency

All 192 generations

Text Rendering87%

Typography Style100%

Writing accuracy83%

Spatial Reasoning75%

Attributes Binding100%

Compositionality89%

Counting56%

Negation56%

Relative Position83%

Scale & Proportions67%

Human realism33%

Faces & Expressions83%

Full Body8%

Hands0%

Multi-Subject50%

Truthfulness41%

Photorealism67%

Physics & Reflections42%

World Knowledge33%

Professional Studio74%

Camera & Lighting83%

Color Precision75%

Photorealism33%

Graphical design79%

Data Visualisation33%

Layout & Design78%

Style Diversity92%