Image Generation on PRISM-Bench
89.3ImaginationCLVR
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| CLVRModel Version=9B2026.05 | 89.3 | 73.4 | 67.6 | 87 | 88.2 | 94 | 75 | 82.1 | |
| Gemini 2.5-Flash-Image2026.05 | 88.6 | 84.2 | 78.5 | 88.6 | 84.5 | 90.5 | 81.1 | 85.3 | |
| CLVRModel Version=4B2026.05 | 87.7 | 64.3 | 54.2 | 83.8 | 86 | 88.3 | 69.9 | 76.3 | |
| NextFlow-RLModel Type=AR Models, Evaluator=GPT-4.12026.01 | 87.1 | 79.3 | 49.8 | 83.8 | 88.8 | 90.5 | 72.4 | 78.8 | |
| GPT-4o2026.05 | 86.4 | 88.2 | 69.7 | 90.7 | 92.1 | 92.8 | 78.3 | 86.3 | |
| NextFlowModel Type=AR Models, Evaluator=GPT-4.12026.01 | 84.3 | 71.2 | 50.3 | 80.2 | 82.8 | 71.2 | 70.8 | 74.7 | |
| Qwen-ImageModel Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 79.6 | 76.3 | 61.6 | 86.6 | 90.4 | 90.3 | 74.5 | 79.9 | |
| Qwen-Image2026.05 | 79.6 | 76.3 | 61.6 | 86.6 | 90.4 | 90.3 | 74.5 | 79.9 | |
| FLUX.2 9B baseModel Version=9B2026.05 | 77.2 | 64.8 | 55.5 | 76.7 | 78.3 | 86.4 | 70 | 72.7 | |
| SEEDream 3.0Model Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 76.9 | 77 | 63.2 | 85.7 | 89.8 | 89.8 | 75 | 79.6 | |
| Seedream 3.02026.05 | 76.9 | 77 | 63.2 | 85.7 | 89.8 | 89.8 | 75 | 79.6 | |
| HiDream-I1-FullModel Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 75 | 73.4 | 64.3 | 83.1 | 89.5 | 87.8 | 57.9 | 75.9 | |
| Uni-CoT2026.05 | 74.4 | 55.5 | 37.6 | 73.7 | 80.9 | 80.1 | 60 | 66.1 | |
| SD3.5-LargeModel Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 72.3 | 74.3 | 58.9 | 80.7 | 86.2 | 85.9 | 58 | 73.7 | |
| SD3.5-Large2026.05 | 72.3 | 74.3 | 58.9 | 80.7 | 86.2 | 85.9 | 58 | 73.9 | |
| FLUX.2 4B baseModel Version=4B2026.05 | 71.8 | 53.1 | 45.5 | 69.5 | 74.1 | 79.2 | 61.6 | 65.7 | |
| Bagel-CoTModel Type=AR + Diffusion Models, Evaluator=GPT-4.12026.01 | 71.3 | 61.2 | 31.7 | 67.3 | 83.8 | 83.2 | 58 | 65.2 | |
| SD3.5-Medium2026.05 | 71.3 | 68.3 | 41.7 | 78.9 | 85.2 | 84.3 | 56.9 | 69.5 | |
| FLUX.1-devModel Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 71.1 | 71 | 56.3 | 76.4 | 89.7 | 86.8 | 64.6 | 73.7 | |
| FLUX.1-dev2026.05 | 71.1 | 71 | 56.3 | 76.4 | 89.7 | 86.8 | 64.6 | 73.9 | |
| HiDream-I1-DevModel Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 69 | 69.5 | 58.8 | 73.7 | 83.7 | 83.7 | 52.8 | 70.2 | |
| BagelModel Type=AR + Diffusion Models, Evaluator=GPT-4.12026.01 | 68.7 | 54.6 | 37.4 | 69.6 | 81.6 | 81.8 | 61.7 | 65.1 | |
| BAGEL2026.05 | 68.7 | 54.6 | 37.4 | 69.6 | 81.6 | 81.8 | 61.7 | 65.1 | |
| JanusPro-7BModel Type=AR Models, Evaluator=GPT-4.12026.01 | 68.1 | 59.5 | 26.1 | 72.6 | 75.4 | 72.4 | 51.1 | 60.7 | |
| Janus-Pro-7B2026.05 | 68.1 | 59.5 | 26.1 | 72.6 | 75.4 | 72.4 | 51.1 | 60.7 | |
| SD3-MediumModel Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 63.3 | 60.6 | 43 | 75.2 | 79.5 | 82.3 | 53.8 | 65.4 | |
| SDXLModel Type=Diffusion Models, Evaluator=GPT-4.12026.01 | 58.2 | 70 | 25.4 | 73.9 | 78 | 75.4 | 41.9 | 60.4 |