Compositional Multi-Image-to-Video Generation on IntelligentVBench 3Subjects with BKG
3.53IFOmniWeaving
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| OmniWeaving#Params=8.3B+7B2026.03 | 3.53 | 4.01 | 4.54 | 3.26 | 4.03 | |
| OmniWeaving#Params=8.3B2026.06 | 3.53 | 4.01 | 4.54 | — | 4.03 | |
| VINO#Params=13B+4B2026.03 | 2.63 | 2.97 | 4.24 | 2.53 | 3.28 | |
| VINO#Params=13B2026.06 | 2.63 | 2.97 | 4.24 | — | 3.28 | |
| VACE-Wan2.1#Params=14B2026.03 | 2.6 | 3.03 | 4.4 | 2.5 | 3.34 | |
| VACE-Wan2.1#Params=14B2026.06 | 2.6 | 3.03 | 4.4 | — | 3.34 | |
| SkyReels-V3#Params=14B2026.03 | 2.59 | 3.1 | 4.3 | 2.37 | 3.33 | |
| SkyReels-V3#Params=14B2026.06 | 2.59 | 3.1 | 4.3 | — | 3.33 | |
| MAGREF#Params=14B2026.03 | 2.5 | 2.21 | 4.46 | 2.07 | 3.06 | |
| LoomVideo (RL)#Params=5B2026.06 | 2.47 | 2.67 | 3.67 | — | 2.94 | |
| LoomVideo (Stage 3)#Params=5B2026.06 | 2.4 | 2.7 | 3.61 | — | 2.9 | |
| Phantom#Params=14B2026.03 | 2.36 | 2.79 | 4.21 | 2.2 | 3.12 | |
| Phantom#Params=14B2026.06 | 2.36 | 2.79 | 4.21 | — | 3.12 | |
| UniVideo(hidden)#Params=13B+7B2026.03 | 2.31 | 2.83 | 3.94 | 2.29 | 3.03 | |
| UniVideo(hidden)#Params=13B2026.06 | 2.31 | 2.83 | 3.94 | — | 3.03 | |
| UniVideo(query)#Params=13B+7B2026.03 | 2.3 | 2.5 | 3.89 | 2.2 | 2.89 | |
| UniVideo(query)#Params=13B2026.06 | 2.3 | 2.5 | 3.89 | — | 2.89 | |
| VACE-LTX#Params=2B2026.03 | 1.94 | 2.06 | 2.41 | 1.89 | 2.14 | |
| VACE-LTX#Params=2B2026.06 | 1.94 | 2.06 | 2.41 | — | 2.14 | |
| SkyReels-A2#Params=14B2026.03 | 1.64 | 1.76 | 2.5 | 1.56 | 1.97 | |
| SkyReels-A2#Params=14B2026.06 | 1.64 | 1.76 | 2.5 | — | 1.97 |