DPG-Bench
Benchmark website →Dense Prompt Graph Benchmark evaluates image generation models on complex, detailed text prompts with multiple requirements.
About this test
- What it measures
- Ability to follow dense, multi-constraint prompts with many simultaneous requirements.
- How it was administered
- Models generate images from long detailed prompts; evaluated on adherence to each constraint; percentage score.
Model rankings
Models ranked by score on this benchmark. Higher is better.
| Rank | Model | Provider | Score | Percentile | Tags |
|---|---|---|---|---|---|
| 1 | Midjourney | 85.0 | p95 | Image Generation, Proprietary | |
| 2 | Black Forest Labs | 84.2 | p93 | Image Generation, Proprietary | |
| 3 |
