Text-to-Video on Video 480x832x97
132.23Inference Time (s)LoomVideo
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LoomVideo#Params=5B + 8B, Source Token Injection=Scale-and-Add2026.06 | 132.23 | 6.24 | |
| Wan 2.2#Params=5B + 5.68B (UMT5-XXL), Source Token Injection=-2026.06 | 138.61 | — | |
| OmniWeaving#Params=8.3B+ 7B, Source Token Injection=Channel Concat2026.06 | 824.93 | — | |
| UniVideo (hidden)#Params=13B+ 7B, Source Token Injection=Token Concat2026.06 | 1,792.65 | — | |
| VINO#Params=13B+ 4B, Source Token Injection=Token Concat2026.06 | 2,793.52 | — |