Text-to-Video Generation on 40 motion-complex prompts (test)
85.5Human AnatomyVeo 3.1
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Veo 3.1Model Type=Closed-source2025.12 | 85.5 | 99.3 | 67.5 | 59.9 | 90.2 | 89.9 | 92.1 | |
| Kling 2.5 TurboModel Type=Closed-source2025.12 | 84.7 | 99 | 70 | 59.7 | 90.4 | 89.3 | 91 | |
| EchoMotionBackbone=Wan-5B2025.12 | 83.2 | 99.1 | 64 | 58.2 | 80.4 | 81.4 | 80.8 | |
| Wan-5BModel Type=Open-source baseline2025.12 | 82.3 | 98.7 | 62.2 | 58.3 | 72.7 | 77.5 | 69.1 |