Spatial Reasoning Generation on T2I-CoReBench (test)
94.5MISeedream 4.0
Evaluation results
| Method | Links | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Seedream 4.0Model Type=Closed-Source Generation Models2026.02 | 94.5 | 88.6 | 79.9 | 95.7 | 89.6 | 79.8 | 53.8 | 60.2 | 89.7 | 84.8 | 80.4 | 74.4 | 45.9 | 71.1 | 77.3 | |
| Nano Banana ProModel Type=Closed-Source Generation Models2026.02 | 91.8 | 81.4 | 82.4 | 98 | 88.4 | 92.9 | 69.8 | 71.7 | 90.8 | 83.7 | 86.9 | 74.9 | 70.8 | 80.2 | 82.9 | |
| SAGenModel Type=MLLM Enhanced Generation Models, Chain-of-Thought=w/ SCoT2026.02 | 89.6 | 79.2 | 73.7 | 86.2 | 82.2 | 91.7 | 64.6 | 66.8 | 87.9 | 77.6 | 80.8 | 71.7 | 68.2 | 76.2 | 78.3 | |
| GPT-Image-1.5Model Type=Closed-Source Generation Models2026.02 | 89.1 | 82.8 | 77.5 | 95.4 | 86.2 | 65.1 | 67.3 | 68 | 92.6 | 76.8 | 87.4 | 78.5 | 54.9 | 73.8 | 77.9 | |
| GPT-ImageModel Type=Closed-Source Generation Models2026.02 | 86.8 | 78.5 | 76.9 | 89 | 82.8 | 62.2 | 55.8 | 62.9 | 88.6 | 70.7 | 83.3 | 72.2 | 50.7 | 68.3 | 73.1 | |
| Nano BananaModel Type=Closed-Source Generation Models2026.02 | 86.5 | 77.4 | 73.3 | 89.8 | 81.8 | 66.9 | 62.3 | 63.1 | 87.8 | 77.6 | 83.8 | 72.8 | 62.2 | 72.1 | 75.3 | |
| Qwen-ImageModel Type=MLLM Enhanced Generation Models2026.02 | 84.9 | 83.2 | 70.7 | 87.4 | 81.5 | 44.7 | 32.5 | 47.3 | 81.9 | 52.4 | 57.4 | 62.8 | 21.5 | 50.1 | 60.5 | |
| FLUX.1-Krea-devModel Type=Diffusion Models2026.02 | 73.2 | 71.1 | 56.4 | 31.2 | 58 | 35.3 | 26.9 | 43.1 | 75.5 | 48.2 | 60.5 | 47.2 | 21.4 | 44.8 | 49.2 | |
| FLUX.1-schnellModel Type=Diffusion Models2026.02 | 67.5 | 64.1 | 52.9 | 23.4 | 51.9 | 30.6 | 24.5 | 38.8 | 69.7 | 42.4 | 59.3 | 41.3 | 16.3 | 40.4 | 44.2 | |
| HiDream-I1Model Type=Diffusion Models2026.02 | 65.2 | 63.9 | 46.9 | 36 | 53 | 36.9 | 24 | 39.4 | 57.5 | 31.7 | 49.2 | 49.2 | 24.7 | 39.1 | 43.7 | |
| show-o2-7BModel Type=MLLM Enhanced Generation Models2026.02 | 63.8 | 62.4 | 50.9 | 31 | 52 | 34.3 | 23.7 | 37 | 57.6 | 40 | 56.9 | 35.3 | 15.2 | 37.5 | 42.4 | |
| FLUX.1-devModel Type=Diffusion Models2026.02 | 61 | 62.3 | 49.8 | 35.8 | 52.2 | 30.2 | 22.6 | 32.5 | 68 | 41.1 | 61.7 | 40.4 | 22.7 | 39.9 | 44 | |
| SD-3.5-MediumModel Type=Diffusion Models2026.02 | 60.5 | 61.8 | 37.8 | 13.4 | 43.4 | 26.9 | 19 | 30 | 56.9 | 30.5 | 53.9 | 36.8 | 16 | 33.7 | 37 | |
| BAGELModel Type=MLLM Enhanced Generation Models, Chain-of-Thought=w/ Think2026.02 | 60.3 | 64.1 | 45 | 3.4 | 43.2 | 32 | 25.5 | 31.9 | 66.5 | 50.3 | 62.1 | 46.5 | 33.3 | 43.5 | 43.4 | |
| SD-3.5-LargeModel Type=Diffusion Models2026.02 | 58.9 | 60.8 | 36.5 | 21.6 | 44.4 | 29.4 | 21.1 | 31 | 58.4 | 32.7 | 56.1 | 42.8 | 18.7 | 36.3 | 39 | |
| BLIP3o-8BModel Type=MLLM Enhanced Generation Models2026.02 | 48.9 | 50.3 | 32.4 | 0.9 | 33.1 | 24.9 | 17.9 | 25.4 | 47 | 39.7 | 54.2 | 40.8 | 15.1 | 33.1 | 33.1 |