Image-to-Video Generation on VBench (CLIP, DINO, Temporal Metrics)
0.9961Motion SmoothnessHunyuanVideo-1.5
Evaluation Results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HunyuanVideo-1.5Steps=12, Params (B)=8.32026.07 | 0.9961 | — | — | — | — | 0.6457 | 0.7078 | 0.9649 | 0.9846 | — | — | 89.83 | 99.39 | 99.49 | |
| Gen3C2026.06 | 0.9955 | — | — | — | — | 0.5673 | 0.725 | 0.9684 | 0.9695 | — | — | — | — | — | |
| MATRIX2025.10 | 0.994 | — | — | — | 0.492 | 0.587 | 0.6973 | 0.956 | 0.962 | — | 0.954 | — | — | — | |
| LTX2.3Steps=12, Params (B)=222026.07 | 0.9938 | — | — | — | — | 0.6306 | 0.7154 | 0.9608 | 0.9839 | — | — | 89.42 | 98.87 | 98.6 | |
| VideoWeave-I2V2026.06 | 0.9936 | — | — | — | — | 0.5483 | 0.7143 | 0.9573 | 0.9684 | — | — | — | — | — | |
| CleanOptimization model=CogVideoX [61], Generation model=LTX-Video [17]2026.07 | 0.9936 | — | — | — | — | 0.6087 | 0.6822 | 0.9815 | 0.977 | 0.9904 | — | — | — | — | |
| CogVideoX-2B-I2VModel Scale=2B2025.10 | 0.993 | — | — | — | 0.152 | 0.602 | 0.6969 | 0.962 | 0.969 | — | 0.937 | — | — | — | |
| Open-Sora-I2V2025.10 | 0.992 | — | — | — | 0.762 | 0.495 | 0.6332 | 0.937 | 0.926 | — | 0.893 | — | — | — | |
| BaselineModel Scale=1.3B2025.10 | 0.992 | — | — | — | 0.378 | 0.498 | — | 0.955 | — | 0.981 | — | — | — | — | |
| Epipolar Geometry OptimizationModel Scale=1.3B2025.10 | 0.992 | — | — | — | 0.343 | 0.499 | — | 0.955 | — | 0.98 | — | — | — | — | |
| TaVid2025.10 | 0.991 | — | — | — | 0.727 | 0.568 | 0.689 | 0.939 | 0.942 | — | 0.919 | — | — | — | |
| Wan2.1-v1.1-FunSteps=20, Params (B)=1.52026.07 | 0.9902 | — | — | — | — | 0.6672 | 0.6955 | 0.9478 | 0.9412 | — | — | 88.81 | 98.18 | 99.31 | |
| LTX-2Steps=60, Params (B)=132026.07 | 0.9901 | — | — | — | — | 0.6286 | 0.6931 | 0.9578 | 0.9821 | — | — | 88.92 | 98.83 | 98.42 | |
| Ours-PruneSteps=20, Params (B)=1.22026.07 | 0.9896 | — | — | — | — | 0.6525 | 0.6835 | 0.9455 | 0.9392 | — | — | 88.34 | 98.1 | 99.22 | |
| ICMDBase Model=Wan2.1-I2V-14B, Active Parameters=11.3B(30%↓), Steps=50×22026.05 | 0.9893 | — | — | — | 12.98 | 0.3072 | 0.4545 | 0.8665 | 0.72 | — | — | 57.79 | — | — | |
| I2VGuardOptimization model=CogVideoX [61], Generation model=LTX-Video [17]2026.07 | 0.989 | — | — | — | — | 0.5578 | 0.6142 | 0.9533 | 0.9452 | 0.9829 | — | — | — | — | |
| Anti-PromptOptimization model=CogVideoX [61], Generation model=LTX-Video [17]2026.07 | 0.9883 | — | — | — | — | 0.554 | 0.6263 | 0.956 | 0.9461 | 0.9814 | — | — | — | — | |
| ViewCrafter2026.06 | 0.9879 | — | — | — | — | 0.5672 | 0.6898 | 0.9717 | 0.9621 | — | — | — | — | — | |
| GeometryForcing2026.06 | 0.987 | — | — | — | — | 0.4875 | 0.6918 | 0.9747 | 0.9653 | — | — | — | — | — | |
| CogVideoX-5B-I2VModel Scale=5B2025.10 | 0.986 | — | — | — | 0.556 | 0.582 | 0.6966 | 0.942 | 0.946 | — | 0.938 | — | — | — | |
| PAREActive Parameters=7.7B(52%↓), Steps=50×22026.05 | 0.9856 | — | — | — | 35.49 | 0.6212 | 0.7057 | 0.9536 | 0.9534 | — | — | 76.24 | — | — | |
| Open-Sora2.0Steps=50, Params (B)=112026.07 | 0.9846 | — | — | — | — | 0.6303 | 0.6846 | 0.9423 | 0.9735 | — | — | 88.54 | 99.48 | 98.76 | |
| I2VGen-XLActive Parameters=1.4B, Steps=50×22026.05 | 0.9843 | — | — | — | 20.28 | 0.6545 | 0.6949 | 0.9727 | 0.9665 | — | — | 74.6 | — | — | |
| PARE (Width-only)Active Parameters=11.3B(30%↓), Steps=50×22026.05 | 0.9839 | — | — | — | 40 | 0.6313 | 0.709 | 0.9525 | 0.9486 | — | — | 77.09 | — | — | |
| CogVideoX1.5Active Parameters=5B, Steps=50×22026.05 | 0.9835 | — | — | — | 34.49 | 0.622 | 0.6944 | 0.9513 | 0.9453 | — | — | 75.69 | — | — | |
| PARE (Routing-only)Active Parameters=10.9B(32%↓), Steps=50×22026.05 | 0.9833 | — | — | — | 36.86 | 0.6507 | 0.6817 | 0.9438 | 0.965 | — | — | 76.55 | — | — | |
| PARE + Step DistillActive Parameters=8.9B(45%↓), Steps=42026.05 | 0.9832 | — | — | — | 33.84 | 0.6297 | 0.6886 | 0.9598 | 0.956 | — | — | 75.93 | — | — | |
| I2VGuardOptimization model=LTX-Video [17], Generation model=CogVideoX [61]2026.07 | 0.9824 | — | — | — | — | 0.6077 | 0.7043 | 0.9862 | 0.9749 | 0.9698 | — | — | — | — | |
| Wan2.1-I2VMode=Teacher, Active Parameters=16.2B, Steps=50×22026.05 | 0.9819 | — | — | — | 42.54 | 0.6503 | 0.7047 | 0.9605 | 0.9523 | — | — | 77.92 | — | — | |
| CleanOptimization model=LTX-Video [17], Generation model=CogVideoX [61]2026.07 | 0.9815 | — | — | — | — | 0.6148 | 0.7023 | 0.9887 | 0.9715 | 0.9699 | — | — | — | — | |
| FFGo2025.11 | 0.98 | 0.67 | 0.46 | 34 | 14.64 | 0.85 | 0.73 | — | — | — | — | — | — | — | |
| AccVideoBase Model=Wan2.1-I2V-14B, Active Parameters=16.2B, Steps=52026.05 | 0.9788 | — | — | — | 18.21 | 0.6309 | 0.6928 | 0.9602 | 0.9562 | — | — | 73.35 | — | — | |
| DMD2Inference steps=4, Base Model=Wan2.2-I2V-A14B2025.10 | 0.9785 | — | — | — | — | 0.6414 | — | 0.9444 | 0.9184 | 0.9573 | — | — | — | — | |
| Anti-PromptOptimization model=LTX-Video [17], Generation model=CogVideoX [61]2026.07 | 0.9783 | — | — | — | — | 0.5956 | 0.6842 | 0.9672 | 0.9528 | 0.9656 | — | — | — | — | |
| Base modelInference steps=40, Base Model=Wan2.2-I2V-A14B2025.10 | 0.9774 | — | — | — | — | 0.6271 | — | 0.9375 | 0.9039 | 0.9552 | — | — | — | — | |
| Phased DMDInference steps=4, Base Model=Wan2.2-I2V-A14B2025.10 | 0.9766 | — | — | — | — | 0.6391 | — | 0.9416 | 0.914 | 0.9517 | — | — | — | — | |
| DMD2Base Model=Wan2.1-I2V-14B, Active Parameters=16.2B, Steps=42026.05 | 0.9747 | — | — | — | 38.59 | 0.6374 | 0.6964 | 0.95 | 0.94 | — | — | 76.4 | — | — | |
| DynamiCrafterActive Parameters=1.4B, Steps=50×22026.05 | 0.9746 | — | — | — | 31.97 | 0.6638 | 0.6909 | 0.9537 | 0.9539 | — | — | 75.94 | — | — | |
| Step-Video-TI2VActive Parameters=30B, Steps=50×22026.05 | 0.974 | — | — | — | 37.34 | 0.6374 | 0.6838 | 0.9612 | 0.9536 | — | — | 76.39 | — | — | |
| CausVidBase Model=Wan2.1-I2V-14B, Active Parameters=16.2B, Steps=42026.05 | 0.9735 | — | — | — | 38.12 | 0.6487 | 0.7085 | 0.9503 | 0.9432 | — | — | 76.76 | — | — | |
| NeoDragonBase Model=Wan2.1-I2V-14B, Active Parameters=12.2B(25%↓), Steps=50×22026.05 | 0.9702 | — | — | — | 30.23 | 0.5912 | 0.674 | 0.9467 | 0.9345 | — | — | 73.65 | — | — | |
| VACE2025.11 | 0.97 | 0.68 | 0.46 | 33.6 | 7.78 | 0.92 | 0.65 | — | — | — | — | — | — | — | |
| SEINEActive Parameters=0.9B, Steps=250×22026.05 | 0.9665 | — | — | — | 25.63 | 0.5921 | 0.7055 | 0.9575 | 0.9502 | — | — | 73.8 | — | — | |
| Wan2.2-I2V-A14B2025.11 | 0.96 | 0.66 | 0.42 | 33.2 | 9.75 | 0.82 | 0.61 | — | — | — | — | — | — | — | |
| SkyReels-A22025.11 | 0.96 | 0.66 | 0.43 | 33.1 | 12.93 | 0.91 | 0.72 | — | — | — | — | — | — | — | |
| F3-PruningBase Model=Wan2.1-I2V-14B, Active Parameters=12.2B(25%↓), Steps=50×22026.05 | 0.9397 | — | — | — | 55.67 | 0.357 | 0.6305 | 0.85 | 0.7193 | — | — | 67.55 | — | — |