Video Generation on VBench (test)
83.4Semantic ScoreERNIE 5.0
Evaluation Results
| Method | Links | ||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ERNIE 5.0Setting=Instruction-tuned2026.02 | 83.4 | — | — | — | — | — | 84.4 | — | — | — | — | — | — | — | — | — | 84.2 | — | — | — | — | — | — | — | — | — | |
| Veo32026.02 | 82.49 | — | — | — | — | — | 85.7 | — | — | — | — | — | — | — | — | — | 85.06 | — | — | — | — | — | — | — | — | — | |
| ERNIE 5.0-BaseSetting=Native multimodal foundation / Base2026.02 | 82.31 | — | — | — | — | — | 84.14 | — | — | — | — | — | — | — | — | — | 83.78 | — | — | — | — | — | — | — | — | — | |
| Sparse Forcingtype=Chunk-wise autoregressive model, Pre-training=with pretraining, Block size=[3,4,4], #Params=1.3B, Resolution=896×512, Throughput (FPS)=19.9, Latency (s)=0.592026.04 | 81.39 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 84.14 | — | 84.84 | — | — | — | — | — | — | |
| Sparse Forcingtype=Chunk-wise autoregressive model, Pre-training=without pretraining, Block size=[3,4,4], #Params=1.3B, Resolution=896×512, Throughput (FPS)=19.9, Latency (s)=0.592026.04 | 81.36 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.99 | — | 84.65 | — | — | — | — | — | — | |
| Self Forcing#Params=1.3B, Resolution=832×480, Denoising Steps=42025.11 | 81.28 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 84.31 | — | 85.07 | — | — | — | — | — | — | |
| ASD (Ours)#Params=1.3B, Resolution=832×480, Denoising Steps=4, FFE strategy=false2025.11 | 81.25 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 84.38 | — | 85.16 | — | — | — | — | — | — | |
| ASD (Ours)#Params=1.3B, Resolution=832×480, Denoising Steps=1, FFE strategy=true2025.11 | 81.24 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.89 | — | 84.55 | — | — | — | — | — | — | |
| Sparse Forcingtype=Chunk-wise autoregressive model, Pre-training=without pretraining, Block size=[1,8,8], #Params=1.3B, Resolution=896×512, Throughput (FPS)=18.8, Latency (s)=0.632026.04 | 81.24 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.91 | — | 84.58 | — | — | — | — | — | — | |
| ASD (Ours)#Params=1.3B, Resolution=832×480, Denoising Steps=2, FFE strategy=true2025.11 | 81.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 84.32 | — | 85.15 | — | — | — | — | — | — | |
| Self Forcingtype=Chunk-wise autoregressive model, #Params=1.3B, Resolution=896×512, Throughput (FPS)=17.0, Latency (s)=0.692026.04 | 81.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.88 | — | 84.6 | — | — | — | — | — | — | |
| SANA-Video#Params=2B, Model Type=Linear Diffusion2026.05 | 80.76 | 97.09 | 96.95 | 89.22 | 95.1 | — | 70.48 | 67.55 | — | — | — | — | — | — | — | — | — | — | — | 86.07 | — | — | — | — | — | — | |
| Self Forcing (Retrained)#Params=1.3B, Resolution=832×480, Denoising Steps=2, Retrained=true2025.11 | 80.62 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.49 | — | 84.2 | — | — | — | — | — | — | |
| Wan2.1#Params=1.3B, Resolution=832×480, Denoising Steps=502025.11 | 80.09 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 84.26 | — | 85.3 | — | — | — | — | — | — | |
| Wan2.1type=Diffusion model, #Params=1.3B, Resolution=832×480, Throughput (FPS)=0.78, Latency (s)=1032026.04 | 80.09 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 84.26 | — | 85.3 | — | — | — | — | — | — | |
| Wan-MHLA-HBackbone=Wan2.1-1.3B, Attention=Hybrid MHLA (2/3 layers)2026.01 | 79.59 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.82 | — | 84.87 | 103 | — | — | — | — | — | |
| NOVA#Params=0.6B, Resolution=768×480, Denoising Steps=252025.11 | 79.05 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80.12 | — | 80.39 | — | — | — | — | — | — | |
| CausVidtype=Chunk-wise autoregressive model, #Params=1.3B, Resolution=896×512, Throughput (FPS)=17.0, Latency (s)=0.692026.04 | 78.49 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 82.69 | — | 83.73 | — | — | — | — | — | — | |
| Self Forcing (Retrained)#Params=1.3B, Resolution=832×480, Denoising Steps=1, Retrained=true2025.11 | 78.35 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80.62 | — | 81.19 | — | — | — | — | — | — | |
| HunyuanVideo-12026.02 | 76.88 | — | — | — | — | — | 85.07 | — | — | — | — | — | — | — | — | — | 83.43 | — | — | — | — | — | — | — | — | — | |
| Wan-MHLABackbone=Wan2.1-1.3B, Attention=MHLA (all layers)2026.01 | 76.16 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 82.62 | — | 84.26 | 81 | — | — | — | — | — | |
| Wan2.1-14B-0725Parameters=14B2026.02 | 76.11 | — | — | — | — | — | 85.59 | — | — | — | — | — | — | — | — | — | 83.69 | — | — | — | — | — | — | — | — | — | |
| Wan-FABackbone=Wan2.1-1.3B, Attention=FlashAttention2026.01 | 75.65 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.31 | — | 85.23 | 166 | — | — | — | — | — | |
| DG+CFGModel=Wan14B2026.07 | 75.06 | 94.36 | 97.62 | — | — | — | 68.94 | 67.81 | 89.52 | — | — | — | — | — | 22.46 | — | — | 83.05 | — | 85.04 | — | — | — | — | — | — | |
| DGModel=Wan14B2026.07 | 74.82 | 94.3 | 97.59 | — | — | — | 68.86 | 67.79 | 89.31 | — | — | — | — | — | 22.34 | — | — | 82.93 | — | 84.96 | — | — | — | — | — | — | |
| S2-GuidanceModel=Wan14B2026.07 | 74.65 | 94.21 | 97.56 | — | — | — | 68.78 | 67.77 | 89.08 | — | — | — | — | — | 22.27 | — | — | 82.84 | — | 84.89 | — | — | — | — | — | — | |
| SkyReels-V2#Params=1.3B, Resolution=960×540, Denoising Steps=302025.11 | 74.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 82.67 | — | 84.7 | — | — | — | — | — | — | |
| SkyReels-V2type=Chunk-wise autoregressive model, #Params=1.3B, Resolution=960×540, Throughput (FPS)=0.49, Latency (s)=1122026.04 | 74.53 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 82.67 | — | 84.7 | — | — | — | — | — | — | |
| CFGModel=Wan14B2026.07 | 73.76 | 94.45 | 97.66 | — | — | — | 68.68 | 67.82 | 84.97 | — | — | — | — | — | 22.14 | — | — | 82.65 | — | 84.88 | — | — | — | — | — | — | |
| Original: 50 stepsCFG Distilled=false, Sampling Steps=50, Latency(s)=1155.3, Speed=1.00x, Peak VRAM=99.23GB2026.02 | 73.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.9 | — | 81.5 | — | — | — | — | — | — | |
| MAGI-1#Params=4.5B, Model Type=Autoregressive Diffusion2026.05 | 72.02 | 95.71 | 97.14 | 97.69 | 98.67 | — | 62.44 | 64.84 | — | — | — | — | — | — | — | — | — | — | — | 86.08 | — | — | — | — | — | — | |
| Causal Forcing#Params=1.3B, Model Type=Distilled Autoregressive2026.05 | 70.97 | 96.19 | 96.44 | 96.27 | 94.24 | — | 68.8 | 70.62 | — | — | — | — | — | — | — | — | — | — | — | 87.09 | — | — | — | — | — | — | |
| DisCaCFG Distilled=true, R=0.2, N=2, Latency(s)=152.8, Speed=7.56x, Peak VRAM=97.64GB2026.02 | 70.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.7 | — | 81.9 | — | — | — | — | — | — | |
| LTX-VideoBase model=Wan-1.3B2026.02 | 70.79 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80 | 82.3 | — | — | — | — | — | — | — | |
| LTX-Video#Params=1.9B, Resolution=768×512, Denoising Steps=202025.11 | 70.79 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80 | — | 82.3 | — | — | — | — | — | — | |
| LTX-Videotype=Diffusion model, #Params=1.9B, Resolution=768×512, Throughput (FPS)=8.98, Latency (s)=13.52026.04 | 70.79 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80 | — | 82.3 | — | — | — | — | — | — | |
| Self Forcing#Params=1.3B, Model Type=Distilled Autoregressive2026.05 | 70.71 | 95.62 | 96.4 | 96.82 | 95.59 | — | 66.75 | 69.96 | — | — | — | — | — | — | — | — | — | — | — | 86.86 | — | — | — | — | — | — | |
| Restricted MeanFlowCFG Distilled=true, Sampling Steps=20, Latency(s)=232.4, Speed=4.97x, Peak VRAM=97.21GB2026.02 | 70.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.5 | — | 81.8 | — | — | — | — | — | — | |
| LongLiveBase model=Wan-1.3B2026.02 | 70.38 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.42 | 84.2 | — | — | — | — | — | — | — | |
| DisCaCFG Distilled=true, R=0.2, N=3, Latency(s)=130.7, Speed=8.84x, Peak VRAM=97.64GB2026.02 | 70.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.5 | — | 81.8 | — | — | — | — | — | — | |
| ARL2 (75%)#Params=1.3B, Model Type=Distilled Linear Autoregressive, Replacement Ratio=75%2026.05 | 70.24 | 94.85 | 95.94 | 96.9 | 94.24 | — | 68.44 | 69.87 | — | — | — | — | — | — | — | — | — | — | — | 86.71 | — | — | — | — | — | — | |
| ARL2 (50%)#Params=1.3B, Model Type=Distilled Linear Autoregressive, Replacement Ratio=50%2026.05 | 70.13 | 95.63 | 96.41 | 97.26 | 95.59 | — | 68.29 | 69.87 | — | — | — | — | — | — | — | — | — | — | — | 87.17 | — | — | — | — | — | — | |
| AMDBase model=Wan-1.3B2026.02 | 69.91 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 82.21 | 85.29 | — | — | — | — | — | — | — | |
| CausVidBase model=Wan-1.3B2026.02 | 69.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.2 | 84.05 | — | — | — | — | — | — | — | |
| CausVid#Params=1.3B, Resolution=832×480, Denoising Steps=42025.11 | 69.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.2 | — | 84.05 | — | — | — | — | — | — | |
| Pyramid Flow#Params=2B, Resolution=640×384, Denoising Steps=202025.11 | 69.62 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.72 | — | 84.74 | — | — | — | — | — | — | |
| DisCaCFG Distilled=true, R=0.2, N=4, Latency(s)=97.7, Speed=11.8x, Peak VRAM=97.64GB2026.02 | 69.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 78.8 | — | 81.1 | — | — | — | — | — | — | |
| Wan2.1-T2V-1.3B+VideoJAM (Reimpl.)Backbone=Wan2.1-T2V-1.3B, Fine-tuning Protocol=VideoJAM, Prompt Enhancement=False, Resolution=480 x 832, Frames=81, FPS=162026.07 | 69.08 | 91.51 | 96.01 | 99.13 | 96.05 | 73.88 | — | — | 79.22 | 59.66 | 79.2 | 88.92 | 66.17 | 28.34 | — | — | 78.76 | — | — | 81.18 | — | — | — | — | — | — | |
| Wan2.1-T2V-1.3B (Full Fine-tune)Backbone=Wan2.1-T2V-1.3B, Fine-tuning Protocol=Full Fine-tune, Prompt Enhancement=False, Resolution=480 x 832, Frames=81, FPS=162026.07 | 68.47 | 93.59 | 95.81 | 99.36 | 97.21 | 54.08 | — | — | 79.9 | 58.2 | 78.98 | 88.57 | 63.31 | 28.55 | — | — | 78.71 | — | — | 81.26 | — | — | — | — | — | — | |
| Restricted MeanFlowCFG Distilled=true, Sampling Steps=9, Latency(s)=108.3, Speed=10.7x, Peak VRAM=97.21GB2026.02 | 67.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 78.4 | — | 81 | — | — | — | — | — | — | |
| MAGI-1Base model=Wan-1.3B2026.02 | 67.74 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.18 | 82.04 | — | — | — | — | — | — | — | |
| MAGI-1#Params=4.5B, Resolution=832×480, Denoising Steps=642025.11 | 67.74 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.18 | — | 82.04 | — | — | — | — | — | — | |
| MAGI-1type=Chunk-wise autoregressive model, #Params=4.5B, Resolution=832×480, Throughput (FPS)=0.19, Latency (s)=2822026.04 | 67.74 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.18 | — | 82.04 | — | — | — | — | — | — | |
| CFG++Model=Wan1.3B2026.07 | 67.43 | 96.7 | 93.28 | — | — | — | 59.02 | 69.14 | 70.06 | — | — | — | — | — | 19.75 | — | — | 80.35 | — | 83.58 | — | — | — | — | — | — | |
| CFG Distilled: 50 stepsCFG Distilled=true, Sampling Steps=50, Latency(s)=581.1, Speed=1.99x, Peak VRAM=97.21GB2026.02 | 66.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 77.9 | — | 80.6 | — | — | — | — | — | — | |
| MeanFlowCFG Distilled=true, Sampling Steps=20, Latency(s)=232.7, Speed=4.96x, Peak VRAM=97.21GB2026.02 | 66.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 78.8 | — | 81.8 | — | — | — | — | — | — | |
| DG+CFGModel=Wan1.3B2026.07 | 66.35 | 96.66 | 96.12 | — | — | — | 60.74 | 69.12 | 78.51 | — | — | — | — | — | 21.24 | — | — | 81.16 | — | 84.97 | — | — | — | — | — | — | |
| DGModel=Wan1.3B2026.07 | 66.12 | 96.61 | 95.96 | — | — | — | 60.61 | 68.96 | 78.33 | — | — | — | — | — | 21.06 | — | — | 81.02 | — | 84.86 | — | — | — | — | — | — | |
| S2-GuidanceModel=Wan1.3B2026.07 | 65.7 | 96.57 | 95.8 | — | — | — | 60.52 | 68.19 | 78.09 | — | — | — | — | — | 20.59 | — | — | 80.93 | — | 84.74 | — | — | — | — | — | — | |
| CFG-ZeroModel=Wan1.3B2026.07 | 65.53 | 96.33 | 94.56 | — | — | — | 59.69 | 69.05 | 78.16 | — | — | — | — | — | 20.31 | — | — | 80.71 | — | 84.51 | — | — | — | — | — | — | |
| TeaCacheCFG Distilled=true, l=0.15, Latency(s)=237.6, Speed=5.00x, Peak VRAM=97.70GB2026.02 | 65.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 77.4 | — | 80.3 | — | — | — | — | — | — | |
| Wan2.1-T2V-1.3BBackbone=Wan2.1-T2V-1.3B, Fine-tuning Protocol=None, Prompt Enhancement=False, Resolution=480 x 832, Frames=81, FPS=162026.07 | 65.43 | 91.83 | 94.71 | 99.17 | 96.51 | 65 | — | — | 76.09 | 53.66 | 74.6 | 89.93 | 62.37 | 20.03 | — | — | 76.93 | — | — | 79.81 | — | — | — | — | — | — | |
| TaylorSeerCFG Distilled=true, N=3, O=1, Latency(s)=268.3, Speed=4.31x, Peak VRAM=130.7GB2026.02 | 65.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 77.5 | — | 80.6 | — | — | — | — | — | — | |
| Wan2.1-T2V-1.3B+VPT(Ours)Backbone=Wan2.1-T2V-1.3B, Fine-tuning Protocol=VPT, Prompt Enhancement=False, Resolution=480 x 832, Frames=81, FPS=162026.07 | 64.86 | 92.65 | 97.27 | 98.64 | 97.85 | 70.83 | — | — | 73.43 | 51.92 | 74.8 | 86.99 | 73.35 | 20.13 | — | — | 79.58 | — | — | 83.25 | — | — | — | — | — | — | |
| CFGModel=Wan1.3B2026.07 | 64.16 | 96.53 | 95.46 | — | — | — | 60.52 | 67.65 | 77.06 | — | — | — | — | — | 20.15 | — | — | 80.29 | — | 84.32 | — | — | — | — | — | — | |
| CFG Distilled: 20 stepsCFG Distilled=true, Sampling Steps=20, Latency(s)=234.4, Speed=4.93x, Peak VRAM=97.21GB2026.02 | 64.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 76.5 | — | 79.6 | — | — | — | — | — | — | |
| FORACFG Distilled=true, N=3, Latency(s)=265.7, Speed=4.35x, Peak VRAM=124.6GB2026.02 | 63.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 76.6 | — | 79.7 | — | — | — | — | — | — | |
| TaylorSeerCFG Distilled=true, N=6, O=1, Latency(s)=166, Speed=6.96x, Peak VRAM=130.7GB2026.02 | 63.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 76.7 | — | 79.9 | — | — | — | — | — | — | |
| Wan2.1-T2V#Params=1.3B, Model Type=Bidirectional Diffusion2026.05 | 63.01 | 96.18 | 98.33 | 98.33 | 96.6 | — | 65.35 | 68.64 | — | — | — | — | — | — | — | — | — | — | — | 87.24 | — | — | — | — | — | — | |
| TeaCacheCFG Distilled=true, l=0.4, Latency(s)=125.3, Speed=9.22x, Peak VRAM=97.70GB2026.02 | 62.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 75.4 | — | 78.7 | — | — | — | — | — | — | |
| Δ-DiTCFG Distilled=true, N=5, Latency(s)=306.7, Speed=3.77x, Peak VRAM=97.68GB2026.02 | 60 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 73.3 | — | 76.7 | — | — | — | — | — | — | |
| CFG Distilled: 10 stepsCFG Distilled=true, Sampling Steps=10, Latency(s)=119.7, Speed=9.65x, Peak VRAM=97.21GB2026.02 | 59 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 73.2 | — | 76.8 | — | — | — | — | — | — | |
| STGModel=Wan1.3B2026.07 | 58.19 | 95.03 | 96.04 | — | — | — | 59.03 | 65.59 | 68.2 | — | — | — | — | — | 21.51 | — | — | 78.78 | — | 83.92 | — | — | — | — | — | — | |
| Original: 10 stepsCFG Distilled=false, Sampling Steps=10, Latency(s)=234.7, Speed=4.92x, Peak VRAM=99.23GB2026.02 | 57.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 72.2 | — | 75.8 | — | — | — | — | — | — | |
| FORACFG Distilled=true, N=6, Latency(s)=144.2, Speed=8.01x, Peak VRAM=124.6GB2026.02 | 57.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 72.6 | — | 76.4 | — | — | — | — | — | — | |
| PABCFG Distilled=true, N=8, Latency(s)=178.8, Speed=6.46x, Peak VRAM=121.3GB2026.02 | 56.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 72.1 | — | 76.1 | — | — | — | — | — | — | |
| PABCFG Distilled=true, N=5, Latency(s)=216.5, Speed=5.34x, Peak VRAM=121.3GB2026.02 | 53.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 69.2 | — | 73.1 | — | — | — | — | — | — | |
| LTX-Video#Params=1.9B, Model Type=Bidirectional Diffusion2026.05 | 51.32 | 87.89 | 92 | 98.2 | 97.91 | — | 48.08 | 58.62 | — | — | — | — | — | — | — | — | — | — | — | 80.45 | — | — | — | — | — | — | |
| SkyReels-V2#Params=1.3B, Model Type=Autoregressive Diffusion, Resolution=544x960, Frames=972026.05 | 48.02 | 97.78 | 97.68 | 97.97 | 98.5 | — | 59.66 | 62.29 | — | — | — | — | — | — | — | — | — | — | — | 85.65 | — | — | — | — | — | — | |
| APGModel=Wan1.3B2026.07 | 45.61 | 96.45 | 95.39 | — | — | — | 49.42 | 64.39 | 59.02 | — | — | — | — | — | 20.01 | — | — | 70.83 | — | 77.13 | — | — | — | — | — | — | |
| Δ-DiTCFG Distilled=true, N=8, Latency(s)=253.7, Speed=4.55x, Peak VRAM=97.68GB2026.02 | 42.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 65.2 | — | 70.9 | — | — | — | — | — | — | |
| Wan-LABackbone=Wan2.1-1.3B, Attention=Linear Attention (all layers)2026.01 | 11.38 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 58.24 | — | 69.96 | 82 | — | — | — | — | — | |
| BLASSTBackbone=Wan2.2-14B2026.06 | — | — | 0 | 98.8 | — | 57.1 | 65.7 | 69.3 | 25.1 | — | — | — | — | 88.8 | — | 26.2 | — | — | — | — | — | — | 1.32 | 61.42 | — | — | |
| BLASSTBackbone=Wan2.2-5B2026.06 | — | — | 96.7 | 99.5 | — | 66.7 | 65.1 | 69.5 | 25.5 | — | — | — | — | 90.5 | — | 25 | — | — | — | — | — | — | 1.41 | 74.41 | — | — | |
| CineScaleResolution (T x H x W)=81*1920*10882026.05 | — | 93.19 | 94.76 | — | 98.12 | 54 | 63.26 | 68.67 | — | — | — | — | — | — | — | — | — | 78.67 | — | — | 3,138 | — | — | 16 | 0 | 38.74 | |
| CineScale-ProResolution (T x H x W)=81*2880*16322026.05 | — | 92.96 | 94.97 | — | 97.96 | 56 | 64.05 | 69.58 | — | — | — | — | — | — | — | — | — | 79.25 | — | — | 15,235 | — | — | 16 | 0 | 188.09 | |
| Cog Video2023.09 | — | 0.9219 | 0.9542 | 0.9764 | 0.9647 | 0.4222 | 0.3818 | 0.4103 | 0.734 | 0.1811 | 0.782 | 0.7957 | 0.1824 | 0.2824 | 0.2201 | 0.078 | 0.077 | — | — | — | — | — | — | — | — | — | |
| CogVideoXPrompt Extension (PE)=with2026.04 | — | — | — | — | 98.5 | 44.5 | 54 | 68.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CogVideoXPrompt Extension (PE)=without2026.04 | — | — | — | — | 98.1 | 45 | 53.6 | 67.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CT-1Prompt Extension (PE)=without2026.04 | — | — | — | — | 99 | 83 | 58.5 | 70.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DenseBackbone=Wan2.1-1.3B, Resolution=480P2026.05 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 195 | 77.15 | 1 | — | — | — | |
| DenseBackbone=Wan2.1-14B, Resolution=480P2026.05 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 1,030 | 79.31 | 1 | — | — | — | |
| DynSparseBackbone=Wan2.2-14B, block size=1282026.06 | — | — | 96.1 | 98.1 | — | 78.6 | 69.7 | 66.5 | 27.6 | — | — | — | — | 87.9 | — | 25.3 | — | — | — | — | — | — | 1.61 | 82.14 | — | — | |
| DynSparseBackbone=Wan2.2-5B, block size=322026.06 | — | — | 97.2 | 99.3 | — | 69.5 | 65.2 | 66.2 | 28.3 | — | — | — | — | 90.5 | — | 24.8 | — | — | — | — | — | — | 1.61 | 81.38 | — | — | |
| FlashAttention3Backbone=Wan2.2-14B2026.06 | — | — | 97.6 | 98.6 | — | 80.7 | 69.5 | 70.9 | 27.8 | — | — | — | — | 91.1 | — | 25.5 | — | — | — | — | — | — | 1 | 0 | — | — | |
| FlashAttention3Backbone=Wan2.2-5B2026.06 | — | — | 96.6 | 99.4 | — | 63.8 | 66.6 | 70 | 27.9 | — | — | — | — | 93.6 | — | 25.8 | — | — | — | — | — | — | 1 | 0 | — | — | |
| FlashVideoResolution (T x H x W)=49*1920*10722026.05 | — | 94.24 | 94.07 | — | 97.57 | 51 | 61.01 | 71.05 | — | — | — | — | — | — | — | — | — | 78.16 | — | — | 124 | — | — | 8 | 0 | 2.53 | |
| Hunyuan2025.11 | — | 95.92 | 96.45 | 99.12 | 98.87 | — | — | 64.61 | — | 69.51 | — | — | 67.55 | — | 19.95 | 24.18 | — | — | — | — | — | — | — | — | — | — | |
| HY-World 1.52026.01 | — | — | — | 97.73 | 98.97 | 72.17 | 54.87 | 65.12 | — | — | — | — | — | — | — | — | 20.16 | — | — | — | — | — | — | — | — | — | |
| LaVie2023.09 | — | 0.9141 | 0.9747 | 0.983 | 0.9638 | 0.4972 | 0.5494 | 0.619 | 0.9182 | 0.3332 | 0.968 | 0.8633 | 0.3409 | 0.5269 | 0.2356 | 0.2593 | 0.2641 | — | — | — | — | — | — | — | — | — |