Video Reasoning on RULER-Bench
68.2Average ScoreVBVR (Wan2.2-5B-Distilled)
Evaluation Results
| Method | Links | |||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| VBVR (Wan2.2-5B-Distilled)Cost (s)=882026.06 | 68.2 | 78.6 | 79.8 | 75.3 | 85.6 | 72.8 | 93.8 | 91 | 63.7 | 85.9 | 79.2 | 83.4 | 57.4 | 70.2 | 53 | 59.6 | 49.5 | 66.9 | 76.6 | 81.5 | 86.7 | 65 | 66.1 | 61.3 | 64.8 | 46.9 | 54 | 51.3 | 51.6 | 56 | 69.1 | |
| Veo 3.1 + PEstrategy=PE2026.06 | 66.2 | 78.5 | 82.4 | 75.1 | 85.5 | 90 | 94.2 | 93.4 | 59.5 | 86.6 | 74.5 | 81.2 | 64.4 | 61.3 | 68.1 | 83.7 | 75.4 | 81.9 | 83.8 | 79.7 | 86.6 | 38.4 | 62.5 | 26.4 | 36.9 | 28.1 | 50.8 | 46.4 | 47.5 | 51.3 | 57.9 | |
| Veo 3.12026.06 | 65 | 80 | 78.1 | 74.9 | 82.8 | 80.8 | 90.9 | 90.9 | 69.2 | 81.1 | 79.1 | 83.3 | 57.3 | 61.5 | 63.4 | 74.5 | 74.1 | 81.3 | 76.5 | 81.4 | 87.3 | 42.2 | 63.9 | 20.1 | 32.5 | 37.5 | 46.9 | 47.9 | 47.5 | 51.3 | 55.3 | |
| Sora 2 + PEstrategy=PE2026.06 | 62.9 | 69.4 | 79.7 | 75.2 | 87.9 | 89.7 | 93.7 | 85.7 | 55.6 | 81.6 | 82.3 | 76.4 | 51.1 | 72.9 | 80.6 | 73.5 | 73 | 84.3 | 75.1 | 82.1 | 79 | 35.6 | 38.2 | 38.2 | 28.8 | 42.5 | 40.3 | 31.3 | 46.9 | 32.5 | 53.7 | |
| Sora 22026.06 | 59.8 | 71.6 | 73.5 | 77.8 | 80.1 | 84.4 | 89.8 | 88.9 | 63.6 | 85.8 | 76.4 | 84.1 | 52.9 | 70.1 | 64.6 | 65.9 | 62.7 | 73.3 | 72.1 | 70.4 | 80.1 | 32.8 | 22.9 | 34 | 30 | 41.9 | 39.6 | 38 | 43.8 | 31.3 | 41.8 | |
| Wan2.2-5B + VideoTPOCost (s)=311, strategy=VideoTPO2026.06 | 50.6 | 49.4 | 62 | 48.6 | 61.7 | 73 | 86.6 | 64.9 | 36.2 | 53.3 | 53.5 | 49.9 | 50.5 | 44.2 | 53.5 | 59.8 | 50.6 | 67.1 | 60.5 | 71.4 | 61.3 | 29.1 | 34.4 | 41.4 | 43.8 | 34.3 | 38.9 | 38.1 | 33.4 | 39.6 | 54.7 | |
| Wan2.2-5B-Distilled + VideoTPOCost (s)=71, strategy=VideoTPO2026.06 | 50.3 | 48.5 | 61.5 | 47.9 | 61.2 | 72.4 | 85.9 | 64.3 | 35.8 | 52.4 | 52.8 | 49.3 | 49.8 | 43.9 | 52.6 | 59.4 | 50.1 | 66.6 | 60.1 | 71 | 60.8 | 29.1 | 34.7 | 41.7 | 43.9 | 34.6 | 39.3 | 38.4 | 33.8 | 39.8 | 55.3 | |
| Wan2.2-5B + Pass@5Cost (s)=490, sampling=Pass@52026.06 | 49.6 | 51.1 | 59.5 | 50.4 | 58.7 | 66 | 83.3 | 60.6 | 41.7 | 51.6 | 53.7 | 54.2 | 48.5 | 44.3 | 45.7 | 53.3 | 42.6 | 58.7 | 54.4 | 68.9 | 63.2 | 32.5 | 36.2 | 37.7 | 45.9 | 40.5 | 40.2 | 44.5 | 35.2 | 41.2 | 50.2 | |
| Wan2.2-14BCost (s)=2122026.06 | 49.4 | 50.7 | 59.7 | 51.7 | 59.4 | 68 | 85.7 | 61 | 40.1 | 51.6 | 54.2 | 52.4 | 48.9 | 43.3 | 47.8 | 55.6 | 44.9 | 61.7 | 57.8 | 70.5 | 64 | 28.4 | 33.3 | 38.9 | 44.4 | 36.9 | 37.5 | 41.7 | 32.5 | 38.1 | 50.2 | |
| Wan2.2-5B-Distilled + Pass@5Cost (s)=90, sampling=Pass@52026.06 | 49.1 | 49.9 | 58.8 | 49.4 | 58 | 65.3 | 82.4 | 59.9 | 41.1 | 50.5 | 52.8 | 53.4 | 47.6 | 43.8 | 44.7 | 52.8 | 42 | 58.1 | 53.9 | 68.3 | 62.5 | 32.1 | 36.2 | 37.8 | 45.6 | 40.4 | 40.3 | 44.4 | 35.3 | 41 | 50.6 | |
| Wan2.2-5B + PECost (s)=101, strategy=PE2026.06 | 48.6 | 48.1 | 60 | 47.4 | 59.5 | 70.2 | 84.5 | 62.6 | 34.2 | 51.2 | 51.7 | 48.7 | 48.5 | 42.7 | 50.4 | 57 | 47.2 | 63.7 | 57.7 | 69.2 | 60.2 | 27.7 | 32.7 | 39.2 | 42.2 | 33.2 | 37.2 | 36.9 | 31.7 | 37.7 | 52.2 | |
| Wan2.2-5B-Distilled + PECost (s)=20, strategy=PE2026.06 | 48.3 | 47.2 | 59.5 | 46.7 | 59 | 69.6 | 83.8 | 62 | 33.8 | 50.3 | 51 | 48.1 | 47.8 | 42.4 | 49.5 | 56.6 | 46.7 | 63.2 | 57.3 | 68.8 | 59.7 | 27.7 | 33 | 39.5 | 42.3 | 33.5 | 37.6 | 37.2 | 32.1 | 37.9 | 52.8 | |
| Wan2.2-5BCost (s)=982026.06 | 46.7 | 48.6 | 57.5 | 47.9 | 56.5 | 65.2 | 81.7 | 59.1 | 37.7 | 49.2 | 50.7 | 50.2 | 46 | 41.5 | 44.9 | 52.5 | 42.2 | 58.2 | 53.7 | 66.7 | 60.7 | 26.7 | 31.2 | 35.7 | 40.7 | 34.7 | 35.2 | 38.7 | 30.2 | 35.7 | 48.2 | |
| Wan2.2-5B-DistilledCost (s)=182026.06 | 46.4 | 47.7 | 57 | 47.2 | 56 | 64.6 | 81 | 58.5 | 37.3 | 48.3 | 50 | 49.6 | 45.3 | 41.2 | 44 | 52.1 | 41.7 | 57.7 | 53.3 | 66.3 | 60.2 | 26.7 | 31.5 | 36 | 40.8 | 35 | 35.6 | 39 | 30.6 | 35.9 | 48.8 |