Image Editing on PhyEditBench anti-physics 1.0
7.07Overall ScoreGemini-2.5
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Gemini-2.5Source Category=closed-source2026.06 | 7.07 | 8.71 | 7.03 | 5.91 | 8.54 | |
| GPT-Image-1.5Source Category=closed-source2026.06 | 7.04 | 9.03 | 7.2 | 5.51 | 8.71 | |
| Seedream4.0Source Category=closed-source2026.06 | 6.95 | 8.2 | 7.29 | 5.78 | 8.11 | |
| PhyWorldSource Category=video-based2026.06 | 6.39 | 7.91 | 6.34 | 5.29 | 7.86 | |
| Qwen-Image-EditSource Category=open-source2026.06 | 6.16 | 7.74 | 6.29 | 4.83 | 7.91 | |
| BAGEL-ThinkSource Category=open-source2026.06 | 6.16 | 7.89 | 6.34 | 5 | 6.83 | |
| FLUX.1-Kontext-devSource Category=open-source2026.06 | 6.07 | 8.6 | 5.71 | 4.43 | 8.63 | |
| UniWorld-V2Source Category=open-source2026.06 | 5.79 | 8.17 | 5.77 | 4.06 | 8.06 | |
| Step1X-EditSource Category=open-source2026.06 | 5.34 | 8.23 | 4.83 | 3.51 | 8.4 | |
| BAGELSource Category=open-source2026.06 | 5.11 | 7.77 | 4.86 | 3.29 | 7.83 | |
| ChronoEdit-14BSource Category=video-based2026.06 | 4.99 | 6.89 | 4.66 | 3.63 | 7.69 | |
| OmniGen2Source Category=open-source2026.06 | 4.74 | 8.17 | 4.09 | 2.89 | 7.23 | |
| InstructPix2PixSource Category=open-source2026.06 | 4.45 | 7.6 | 3.69 | 2.54 | 8.09 | |
| Frame2FrameSource Category=video-based2026.06 | 3.81 | 6.8 | 2.51 | 2.51 | 6.89 |