Video Reasoning on FlowFree (test)
44.3PrecisionVideoRLVR
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| VideoRLVRModel Category=RL Model, Training Epochs=1 (from SFT Epoch 5)2026.05 | 44.3 | 43.8 | 44 | 7.9 | |
| SFT Epoch 10Model Category=SFT Models, Training Epochs=102026.05 | 43.1 | 42.5 | 42.8 | 2.5 | |
| SFT Epoch 5Model Category=SFT Models, Training Epochs=52026.05 | 42.8 | 42.2 | 42.4 | 2.4 | |
| Veo 3.1Model Category=Proprietary2026.05 | 23.9 | 4.7 | 7.5 | 4 | |
| Wan-R1Model Category=SFT Models2026.05 | 20.9 | 3.6 | 6.1 | 0 | |
| Kling V3Model Category=Proprietary2026.05 | 18.8 | 2.7 | 4.7 | 0 | |
| CogVideoX1.5Model Category=Open-Source2026.05 | 18.7 | 2.2 | 3.9 | 0 | |
| VBVR-Wan2.2Model Category=SFT Models2026.05 | 17.9 | 5.6 | 8.5 | 1.7 | |
| Wan2.2-TI2V-5BModel Category=Open-Source2026.05 | 17.4 | 2 | 3.4 | 0 | |
| HunyuanVideoModel Category=Open-Source2026.05 | 12.5 | 2.9 | 4.8 | 0 | |
| Sora 2Model Category=Proprietary2026.05 | 10.8 | 5.1 | 5.8 | 0 |