Text-to-Video Generation on VideoPhy
40.1PC ScoreVideoREPA
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| VideoREPADetailed Prompts=true2026.04 | 40.1 | 72.1 | — | — | — | — | — | — | — | — | |
| PhantomDetailed Prompts=true2026.04 | 39.4 | 70.3 | — | — | — | — | — | — | — | — | |
| PhantomDetailed Prompts=false2026.04 | 37.9 | 47.5 | — | — | — | — | — | — | — | — | |
| PhyT2V (Round 4)Detailed Prompts=false2026.04 | 37 | 61 | — | — | — | — | — | — | — | — | |
| WISADetailed Prompts=false2026.04 | 33 | 62 | — | — | — | — | — | — | — | — | |
| LaVIEDetailed Prompts=false2026.04 | 31.5 | 48.7 | — | — | — | — | — | — | — | — | |
| CogVideoX-5BDetailed Prompts=false2026.04 | 31.4 | 63.1 | — | — | — | — | — | — | — | — | |
| VideoCrafter2Detailed Prompts=false2026.04 | 29.7 | 50.3 | — | — | — | — | — | — | — | — | |
| Wan2.2-TI2V-5BDetailed Prompts=true2026.04 | 28.6 | 64.7 | — | — | — | — | — | — | — | — | |
| Wan2.2-TI2V-5BDetailed Prompts=false2026.04 | 25.2 | 41.5 | — | — | — | — | — | — | — | — | |
| VideoREPADetailed Prompts=false2026.04 | 22.4 | 51.9 | — | — | — | — | — | — | — | — | |
| Cosmos-Diffusion-7BDetailed Prompts=false2026.04 | 18 | 57 | — | — | — | — | — | — | — | — | |
| CogVideoX-5BRound=42024.11 | 0.42 | 0.59 | — | — | — | — | — | — | — | — | |
| CogVideoX-5BRound=32024.11 | 0.39 | 0.56 | — | — | — | — | — | — | — | — | |
| PhantomWan2.2Backbone=Wan2.2-TI2V-5B2026.04 | 0.379 | 0.475 | — | — | — | — | — | — | — | — | |
| PhyT2VSetting=Round 42026.04 | 0.37 | 0.61 | — | — | — | — | — | — | — | — | |
| VideoCrafterRound=42024.11 | 0.33 | 0.49 | — | — | — | — | — | — | — | — | |
| ChatGPT 4T2V Model=CogVideoX-5B2024.11 | 0.33 | 0.41 | — | — | — | — | — | — | — | — | |
| WISA2026.04 | 0.33 | 0.62 | — | — | — | — | — | — | — | — | |
| CogVideoX-5BRound=22024.11 | 0.32 | 0.52 | — | — | — | — | — | — | — | — | |
| LaVIE2026.04 | 0.315 | 0.487 | — | — | — | — | — | — | — | — | |
| CogVideoX-5B2026.04 | 0.314 | 0.631 | — | — | — | — | — | — | — | — | |
| OpenSoraRound=42024.11 | 0.31 | 0.47 | — | — | — | — | — | — | — | — | |
| VideoCrafter22026.04 | 0.297 | 0.503 | — | — | — | — | — | — | — | — | |
| CogVideoX-2BRound=42024.11 | 0.29 | 0.42 | — | — | — | — | — | — | — | — | |
| OpenSoraRound=22024.11 | 0.29 | 0.38 | — | — | — | — | — | — | — | — | |
| VideoCrafterRound=32024.11 | 0.29 | 0.44 | — | — | — | — | — | — | — | — | |
| CogVideoX-2BRound=32024.11 | 0.27 | 0.4 | — | — | — | — | — | — | — | — | |
| OpenSoraRound=32024.11 | 0.27 | 0.44 | — | — | — | — | — | — | — | — | |
| CogVideoX-5BRound=12024.11 | 0.26 | 0.48 | — | — | — | — | — | — | — | — | |
| Wan2.2-TI2V-5B2026.04 | 0.252 | 0.415 | — | — | — | — | — | — | — | — | |
| VideoCrafterRound=22024.11 | 0.25 | 0.38 | — | — | — | — | — | — | — | — | |
| PromptistT2V Model=CogVideoX-5B2024.11 | 0.25 | 0.39 | — | — | — | — | — | — | — | — | |
| VideoREPA2026.04 | 0.224 | 0.519 | — | — | — | — | — | — | — | — | |
| ChatGPT 4T2V Model=OpenSora2024.11 | 0.21 | 0.32 | — | — | — | — | — | — | — | — | |
| CogVideoX-2BRound=22024.11 | 0.19 | 0.32 | — | — | — | — | — | — | — | — | |
| PromptistT2V Model=OpenSora2024.11 | 0.19 | 0.33 | — | — | — | — | — | — | — | — | |
| Cosmos-Diffusion-7B2026.04 | 0.18 | 0.57 | — | — | — | — | — | — | — | — | |
| OpenSoraRound=12024.11 | 0.17 | 0.29 | — | — | — | — | — | — | — | — | |
| VideoCrafterRound=12024.11 | 0.15 | 0.24 | — | — | — | — | — | — | — | — | |
| CogVideoX-2BRound=12024.11 | 0.13 | 0.22 | — | — | — | — | — | — | — | — | |
| CogVideoX-2BBackbone=CogVideoX-2B, Distillation strategy=false2026.04 | — | — | — | 0.146 | — | 0.24 | — | 0.258 | 0.46 | 0.204 | |
| CogVideoX-5BBackbone=CogVideoX-5B, Distillation strategy=false2026.04 | — | — | — | 0.201 | — | 0.325 | — | 0.308 | 0.469 | 0.271 | |
| CogVideoX1.5-5BBackbone=CogVideoX1.5-5B, Distillation strategy=false2026.04 | — | — | — | 0.369 | — | 0.474 | — | 0.435 | 0.557 | 0.428 | |
| DreamMachine2026.01 | — | — | 0.551 | 0.217 | 0.596 | 0.233 | 0.582 | 0.182 | 0.575 | 0.218 | |
| DreamWorld2026.02 | — | — | 0.545 | 0.245 | 0.486 | 0.254 | 0.601 | 0.327 | 0.529 | 0.262 | |
| HunyuanVideo2026.01 | — | — | 0.552 | 0.161 | 0.671 | 0.301 | 0.545 | 0.545 | 0.602 | 0.282 | |
| LaVIE2026.01 | — | — | 0.408 | 0.183 | 0.486 | 0.37 | 0.691 | 0.509 | 0.487 | 0.315 | |
| MAGI-12026.01 | — | — | 0.422 | 0.19 | 0.672 | 0.277 | 0.519 | 0.335 | 0.544 | 0.25 | |
| MAGI-1 + WMReward(∇+BoN)Search method=∇+BoN, Number of particles (N)=82026.01 | — | — | 0.294 | 0.287 | 0.616 | 0.301 | 0.4 | 0.527 | 0.448 | 0.331 | |
| MAGI-1 + WMReward(BoN)Search method=BoN, Number of particles (N)=82026.01 | — | — | 0.308 | 0.28 | 0.596 | 0.363 | 0.455 | 0.364 | 0.453 | 0.328 | |
| MMPhysVideoBackbone=CogVideoX-2B, Distillation strategy=true2026.04 | — | — | — | 0.191 | — | 0.295 | — | 0.344 | 0.489 | 0.26 | |
| MMPhysVideoBackbone=CogVideoX-5B, Distillation strategy=true2026.04 | — | — | — | 0.231 | — | 0.387 | — | 0.363 | 0.506 | 0.321 | |
| MMPhysVideoBackbone=Wan2.1-1.3B, Distillation strategy=true2026.04 | — | — | — | 0.371 | — | 0.469 | — | 0.475 | 0.609 | 0.437 | |
| OmniVDiffBackbone=OmniVDiff, Distillation strategy=false2026.04 | — | — | — | 0.117 | — | 0.233 | — | 0.129 | 0.394 | 0.168 | |
| PhyT2VBackbone=CogVideoX-2B, Distillation strategy=false2026.04 | — | — | — | 0.175 | — | 0.274 | — | 0.184 | 0.468 | 0.219 | |
| PhyT2VBackbone=CogVideoX-5B, Distillation strategy=false2026.04 | — | — | — | 0.226 | — | 0.334 | — | 0.306 | 0.472 | 0.285 | |
| VideoCrafter22026.01 | — | — | 0.504 | 0.322 | 0.507 | 0.274 | 0.481 | 0.291 | 0.503 | 0.297 | |
| VideoJAM2026.02 | — | — | 0.517 | 0.294 | 0.384 | 0.233 | 0.473 | 0.201 | 0.453 | 0.253 | |
| VideoREPABackbone=CogVideoX-2B, Distillation strategy=false2026.04 | — | — | — | 0.195 | — | 0.31 | — | 0.214 | 0.485 | 0.243 | |
| VideoREPABackbone=CogVideoX-5B, Distillation strategy=false2026.04 | — | — | — | 0.223 | — | 0.322 | — | 0.391 | 0.495 | 0.299 | |
| vLDM2026.01 | — | — | 0.466 | 0.207 | 0.675 | 0.286 | 0.524 | 0.486 | 0.569 | 0.28 | |
| vLDM + WMReward(∇+BoN)Search method=∇+BoN, Number of particles (N)=82026.01 | — | — | 0.392 | 0.252 | 0.651 | 0.37 | 0.5 | 0.5 | 0.535 | 0.343 | |
| vLDM + WMReward(BoN)Search method=BoN, Number of particles (N)=82026.01 | — | — | 0.427 | 0.287 | 0.596 | 0.363 | 0.385 | 0.5 | 0.503 | 0.349 | |
| Wan2.1-1.3BBackbone=Wan2.1-1.3B, Distillation strategy=false2026.04 | — | — | — | 0.365 | — | 0.439 | — | 0.42 | 0.594 | 0.405 | |
| Wan2.1-T2V-1.3B2026.02 | — | — | 0.511 | 0.223 | 0.452 | 0.191 | 0.454 | 0.236 | 0.477 | 0.212 | |
| Wan2.1-T2V-1.3BFine-tuning=true2026.02 | — | — | 0.462 | 0.182 | 0.478 | 0.226 | 0.509 | 0.236 | 0.451 | 0.209 | |
| Wan2.1-T2V-1.3BModel Backbone=Wan2.1, Parameter Scale=1.3B, Adaptation Method=None2026.07 | — | — | 0.511 | 0.223 | 0.452 | 0.191 | 0.454 | 0.236 | 0.477 | 0.212 | |
| Wan2.1-T2V-1.3B (Full Fine-tune)Model Backbone=Wan2.1, Parameter Scale=1.3B, Adaptation Method=Full Fine-tune2026.07 | — | — | 0.462 | 0.182 | 0.478 | 0.226 | 0.509 | 0.236 | 0.451 | 0.209 | |
| Wan2.1-T2V-1.3B+VideoJAMModel Backbone=Wan2.1, Parameter Scale=1.3B, Adaptation Method=VideoJAM2026.07 | — | — | 0.454 | 0.132 | 0.561 | 0.273 | 0.401 | 0.309 | 0.491 | 0.221 | |
| Wan2.1-T2V-1.3B+VPTModel Backbone=Wan2.1, Parameter Scale=1.3B, Adaptation Method=VPT2026.07 | — | — | 0.692 | 0.237 | 0.63 | 0.24 | 0.691 | 0.301 | 0.665 | 0.25 | |
| Wan2.1-T2V-14BModel Backbone=Wan2.1, Parameter Scale=14B, Adaptation Method=None2026.07 | — | — | 0.468 | 0.181 | 0.657 | 0.253 | 0.545 | 0.309 | 0.561 | 0.232 | |
| Wan2.1-T2V-14B (Full Fine-tune)Model Backbone=Wan2.1, Parameter Scale=14B, Adaptation Method=Full Fine-tune2026.07 | — | — | 0.638 | 0.19 | 0.601 | 0.231 | 0.64 | 0.22 | 0.621 | 0.215 | |
| Wan2.1-T2V-14B+VPTModel Backbone=Wan2.1, Parameter Scale=14B, Adaptation Method=VPT2026.07 | — | — | 0.678 | 0.273 | 0.678 | 0.315 | 0.673 | 0.327 | 0.677 | 0.3 |