Video Generation on VBench 2.0 (test)
83.79Total ScoreWan2.1 1.3B* + ReHyAt
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Wan2.1 1.3B* + ReHyAtArchitecture=Linear/Hybrid Models, Number of hybrid blocks=15, Chunk size (Tc)=3, Overlap size (To)=12026.01 | 83.79 | — | 80.7 | — | — | — | 84.57 | |
| SANA-VideoArchitecture=Linear/Hybrid Models2026.01 | 83.71 | — | 81.35 | — | — | — | 84.35 | |
| Wan2.1 1.3BModel scale=1.3B parameters2026.01 | 83.31 | — | 75.65 | — | — | — | 85.23 | |
| Attention Surgery (15×R2)Architecture=Linear/Hybrid Models2026.01 | 83.21 | — | 75.25 | — | — | — | 85.19 | |
| Wan2.1 1.3BModel scale=1.3B parameters, Evaluation pipeline=Authors Reproduction2026.01 | 83.1 | — | 75.12 | — | — | — | 85.1 | |
| Mobile Video DiT - ServerModel scale=Up to 2B parameters2026.01 | 83.09 | — | 76.86 | — | — | — | 84.65 | |
| STAArchitecture=Linear/Hybrid Models2026.01 | 83 | — | 73.52 | — | — | — | 85.37 | |
| VSAArchitecture=Linear/Hybrid Models2026.01 | 82.77 | — | 79.47 | — | — | — | 83.6 | |
| CogVideoX1.5 5BModel scale=5B parameters2026.01 | 82.01 | — | 79.17 | — | — | — | 82.72 | |
| CogVideoX 5BModel scale=5B parameters2026.01 | 81.91 | — | 77.33 | — | — | — | 83.05 | |
| M4VArchitecture=Linear/Hybrid Models2026.01 | 81.91 | — | 76.1 | — | — | — | 83.36 | |
| PyramidalFlowModel scale=Up to 2B parameters2026.01 | 81.72 | — | 69.62 | — | — | — | 84.74 | |
| NeodragonModel scale=Up to 2B parameters2026.01 | 81.61 | — | 73.36 | — | — | — | 83.68 | |
| CogVideoX 2BModel scale=2B parameters2026.01 | 81.55 | — | 77.81 | — | — | — | 82.48 | |
| Mobile Video DiT - MobileModel scale=Up to 2B parameters2026.01 | 81.45 | — | 74.76 | — | — | — | 83.12 | |
| Hummingbird 16frameModel scale=Up to 2B parameters2026.01 | 81.35 | — | 71.84 | — | — | — | 83.73 | |
| SnapGenVModel scale=Up to 2B parameters2026.01 | 81.14 | — | 71.84 | — | — | — | 83.47 | |
| LTX-VideoModel scale=Up to 2B parameters2026.01 | 80 | — | 70.79 | — | — | — | 82.3 | |
| Open-Sora V1.2Model scale=Up to 2B parameters2026.01 | 79.76 | — | 73.39 | — | — | — | 81.35 | |
| Open-Sora Plan V1.3Model scale=2B-5B parameters2026.01 | 77.23 | — | 65.62 | — | — | — | 80.14 | |
| Efficient VDiTArchitecture=Linear/Hybrid Models2026.01 | 76.14 | — | — | — | — | — | — | |
| Wan-DMDSteps=42026.01 | 57.48 | 50.87 | 58.16 | 36.36 | 82.8 | 59.23 | — | |
| Wan-DMDSteps=22026.01 | 56.67 | 47 | 62.81 | 37.07 | 80.73 | 55.72 | — | |
| ReHyAtArchitecture Configuration=15, Chunk Size (Tc)=52026.01 | 56.3 | 55.7 | 64.2 | 31.9 | 79.8 | 49.7 | — | |
| SARA2026.05 | 56.19 | 55.38 | 61.11 | 30.91 | 85.07 | 48.5 | — | |
| ReHyAtArchitecture Configuration=15, Chunk Size (Tc)=32026.01 | 56.1 | 55.1 | 62.7 | 30.8 | 81.9 | 50 | — | |
| Wan2.1-1.3BSteps=502026.01 | 56.02 | 48.73 | 63.38 | 33.96 | 80.71 | 53.3 | — | |
| Wan2.1Number of Parameters=1.3B2026.01 | 56 | 48.7 | 63.4 | 34 | 80.7 | 53.8 | — | |
| MoAlign2026.05 | 55.81 | 56.75 | 59.67 | 30.08 | 84.75 | 47.82 | — | |
| Wan2.1-1.3BSteps=252026.01 | 55.73 | 49.49 | 62.5 | 35.01 | 79.44 | 52.2 | — | |
| Wan-AdvSteps=42026.01 | 55.4 | 50.96 | 57.33 | 32.56 | 85.58 | 50.57 | — | |
| Diffusion-DRFBackbone=Wan2.1-1.3B-T2V2026.01 | 55.38 | 54.58 | 56.96 | 27.98 | 80.51 | 56.85 | — | |
| PyramidalWan-DMD-OTInference Schedule=2-2-12026.01 | 55.36 | 50.04 | 64.34 | 29.05 | 77.5 | 56.78 | — | |
| VideoREPA2026.05 | 55.24 | 54.08 | 61.11 | 31.54 | 82.78 | 46.67 | — | |
| Attention SurgeryArchitecture Configuration=15×R22026.01 | 55.1 | 47.5 | 63.1 | 33.4 | 78.9 | 52.8 | — | |
| SFT2026.05 | 55.08 | 54.6 | 59.68 | 29.59 | 80.41 | 51.09 | — | |
| Pretrained Wan2.22026.05 | 55 | 52.56 | 58.5 | 30.98 | 86.04 | 46.89 | — | |
| PyramidalWanInference Schedule=20-20-102026.01 | 54.93 | 44.64 | 64.33 | 28.39 | 85.38 | 51.89 | — | |
| Wan-AdvSteps=22026.01 | 54.82 | 47.07 | 58.47 | 31.51 | 82.72 | 54.36 | — | |
| PyramidalWan-Adv-PDInference Schedule=2-2-12026.01 | 54.3 | 45 | 61.98 | 25.85 | 87.65 | 51.01 | — | |
| Wan-PPF-DMDInference Schedule=2-2-12026.01 | 53.45 | 40.04 | 62.84 | 24.76 | 91.61 | 47.99 | — | |
| CogVideoX-1.5Number of Parameters=5B2026.01 | 53.4 | 43.7 | 63.2 | 29.6 | 72.1 | 48.2 | — | |
| Wan-DMDSteps=12026.01 | 53.17 | 38.69 | 59.31 | 35.36 | 77.49 | 55 | — | |
| PyramidalWan-Adv-ODInference Schedule=2-2-12026.01 | 52.29 | 44.8 | 60.86 | 22.86 | 81.74 | 51.17 | — | |
| Diffusion-DRFBackbone=CogVideoX2026.01 | 51.89 | 40.63 | 57.75 | 27.11 | 81.28 | 52.69 | — | |
| PyramidalWan-DMD-PT*Inference Schedule=2-2-12026.01 | 51.75 | 34.81 | 63.18 | 28.48 | 81.38 | 50.92 | — | |
| Wan-AdvSteps=12026.01 | 50.36 | 38.51 | 54.41 | 29.99 | 79.6 | 49.29 | — | |
| CogVideoXBackbone=CogVideoX2026.01 | 49.52 | 37.88 | 54.87 | 24.56 | 78.1 | 52.18 | — | |
| Flow-DPOBackbone=Wan2.1-1.3B-T2V2026.01 | 49.27 | 41.77 | 50.28 | 27.76 | 71.76 | 54.78 | — |