Video Generation on VideoJAM-bench
98.84Motion ScorePredict-and-Perturb
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Predict-and-PerturbBackbone=Wan2.1-1.3B, NFE=74, Time=1.5×2026.01 | 98.84 | 73.31 | 94.95 | 61.43 | |
| Predict-and-PerturbBackbone=Wan2.1-14B, NFE=74, Time=1.5×2026.01 | 98.37 | 77.34 | 94.77 | 63.08 | |
| Wan2.1Backbone=1.3B, NFE=100, Variant=NFE×2, Time=2.0×2026.01 | 98.23 | 77.34 | 94.23 | 61.6 | |
| Wan2.1Backbone=1.3B, NFE=502026.01 | 98.21 | 75 | 94.05 | 61.1 | |
| TeDiOBackbone=Wan2.12026.05 | 98.21 | 70.31 | 94.78 | 59.22 | |
| Wan2.1Backbone=14B, NFE=502026.01 | 98.1 | 77.34 | 94.22 | 61.92 | |
| TeDiOBackbone=CogVideoX2026.05 | 98.06 | 89.84 | 92.15 | 54.01 | |
| Wan2.1Backbone=14B, NFE=100, Variant=NFE×2, Time=2.0×2026.01 | 98.01 | 77.34 | 94.32 | 61.95 | |
| CFG-ZeroBackbone=Wan2.1-1.3B, NFE=50, Time=1.0×2026.01 | 98.01 | 85.16 | 93.71 | 60.71 | |
| CFG-ZeroBackbone=Wan2.1-14B, NFE=50, Time=1.0×2026.01 | 98 | 78.13 | 94.2 | 61.63 | |
| FlowMoBackbone=Wan2.1-1.3B, NFE=50*, Time=3.3×2026.01 | 97.89 | 75 | 93.77 | 59.88 | |
| Wan2.1Backbone=Wan2.12026.05 | 97.59 | 81.25 | 93.1 | 58.35 | |
| FlowMoBackbone=Wan2.1-14B, NFE=50*, Time=3.3×2026.01 | 97.49 | 79.17 | 93.4 | 60.89 | |
| CogVideoXBackbone=CogVideoX2026.05 | 97.4 | 96.88 | 89.85 | 53.7 |