Video Generation on WorldArena
68.2Interaction QualityKVAF-conditioned generation
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| KVAF-conditioned generation2026.05 | 68.2 | 49.4 | 97.7 | 85 | 78.4 | 90.1 | — | 78.13 | |
| CtrlWorld2026.05 | 62.1 | 47.7 | 93 | 79.6 | 72.7 | 89.1 | — | 74.03 | |
| MIND-VFine-tuned=Bridge V22026.03 | 57 | 13.7 | 70.5 | 78.8 | 59.3 | 86.5 | 7.9 | — | |
| RoboStereoModel Identity=Ours2026.03 | 57 | 34.2 | 88.1 | 79.7 | 72.3 | 88.4 | 8.8 | — | |
| IRASim2026.05 | 56.6 | 36.4 | 93.1 | 77.9 | 66 | 88.5 | — | 69.75 | |
| IRASimFine-tuned=Bridge V22026.03 | 56 | 25.7 | 86.2 | 77.4 | 64 | 85.2 | 5.2 | — | |
| Cosmos-Predict 2.5Conditioning=action2026.05 | 55 | 29.5 | 88.6 | 76.4 | 58.4 | 88.8 | — | 66.12 | |
| RoboMaster2026.05 | 53.6 | 11.6 | 83.4 | 75.9 | 57.7 | 87.6 | — | 61.63 | |
| Cosmos 2.5Fine-tuned=Bridge V22026.03 | 53.4 | 28.6 | 85.4 | 75.6 | 60.4 | 85 | 6.3 | — | |
| WoWFine-tuned=Bridge V22026.03 | 53.1 | 20.4 | 72 | 74.2 | 56.5 | 87.3 | 4.2 | — | |
| RoboMasterFine-tuned=Bridge V22026.03 | 52.6 | 11.5 | 80.1 | 73.1 | 56.1 | 83.9 | 7.5 | — | |
| GigaWorld-0Fine-tuned=Bridge V22026.03 | 52.3 | 15.2 | 61.1 | 75.3 | 59.9 | 84.4 | 10.9 | — | |
| VidarFine-tuned=Bridge V22026.03 | 52 | 18.5 | 77.5 | 74.8 | 57.3 | 85.6 | 3.9 | — | |
| Genie EnvisionerFine-tuned=Bridge V22026.03 | 19.8 | 6.7 | 84.4 | 51.3 | 20 | 84.8 | 1.1 | — |