Text-to-Image Generation on DetailMaster
34.97CLIPScoreStableDiffusion-3.5
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| StableDiffusion-3.5Method Category=State-of-the-Art T2I Models2026.04 | 34.97 | 22.37 | 21.63 | 86.12 | 13.39 | |
| PRISM-QwenMethod Category=State-of-the-Art T2I Models, Backbone=Qwen-Image2026.04 | 34.12 | 22.93 | 22.04 | 86.21 | 12.05 | |
| Qwen-ImageMethod Category=State-of-the-Art T2I Models2026.04 | 33.85 | 22.25 | 20.98 | 85.02 | 8.56 | |
| LongAlignMethod Category=Long-Text-to-Image Methods2026.04 | 33.43 | 22.35 | 24.43 | 82.01 | 13.26 | |
| FLUX-Dev.Method Category=State-of-the-Art T2I Models2026.04 | 33.3 | 22.56 | 21.89 | 86.19 | 13.17 | |
| PRISM-SD1.5Method Category=Long-Text-to-Image Methods, Backbone=Stable Diffusion-1.52026.04 | 32.56 | 22.24 | 24.5 | 83.22 | 13.03 | |
| ELLAMethod Category=Long-Text-to-Image Methods2026.04 | 30.89 | 20.34 | 20.72 | 73.3 | 6.78 |