Image Generation on UniGen Bench (test)
84.13ActionQwen-Image
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Qwen-Image2026.02 | 84.13 | 87.61 | 73.32 | 85.52 | 60.29 | 53.64 | 79.7 | 95.1 | 76.14 | 94.3 | 78.81 | |
| Seedream 3.0Version=3.02026.02 | 82.98 | 85.58 | 73.84 | 87.31 | 61.36 | 52.73 | 80.84 | 98.1 | 71.55 | 95.25 | 78.98 | |
| Imagen-3.0Version=3.02026.02 | 81.46 | 77.33 | 71.71 | 81.34 | 69.84 | 48.36 | 82.86 | 89.25 | 21.55 | 94.75 | 71.85 | |
| FLUX-Kontext-Pro2026.02 | 77.66 | 79.2 | 72.68 | 84.47 | 72.69 | 55.68 | 79.34 | 94.78 | 50.29 | 91.61 | 75.84 | |
| FLUX.1 Ultra2026.02 | 76.5 | 76.5 | 67.78 | 81.54 | 70.7 | 43.18 | 77.54 | 90.6 | 37.36 | 91.61 | 70.67 | |
| FAIL-PDVariant=PD2026.02 | 74.81 | 79.59 | 70.62 | 83.02 | 65.11 | 44.04 | 80.2 | 94.5 | 52.87 | 92.25 | 73.7 | |
| FAIL-PGVariant=PG2026.02 | 72.81 | 77.03 | 67.65 | 82.84 | 66.98 | 43.35 | 78.17 | 94.8 | 51.44 | 91.3 | 72.64 | |
| Z-Image-Turbo2026.02 | 69.3 | 74.54 | 63.03 | 78.36 | 64.57 | 39.68 | 71.57 | 90 | 70.69 | 92.25 | 71.4 | |
| Janus-Pro-7BParameters=7B2026.02 | 64.26 | 67.74 | 62.11 | 72.01 | 64.44 | 37.05 | 68.4 | 90.8 | 2.59 | 86.71 | 61.61 | |
| FLUX.1 Dev2026.02 | 62.17 | 67.84 | 47.04 | 71.84 | 60.96 | 30.91 | 67.21 | 83.9 | 32.18 | 88.92 | 61.3 | |
| SD3.5 LargeScale=Large2026.02 | 62.17 | 68.59 | 58.76 | 69.03 | 58.96 | 32.27 | 69.8 | 88.6 | 32.76 | 88.92 | 62.99 | |
| Hunyuan-DiT2026.02 | 49.05 | 62.71 | 41.62 | 44.78 | 55.48 | 24.55 | 59.64 | 94.1 | 1.15 | 80.7 | 51.38 |