Video Understanding on Motion Bench
58AccuracyCapRL-Video-178K
Evaluation Results
| Method | Links | |
|---|---|---|
| CapRL-Video-178KBackbone=Molmo2-8B2026.06 | 58 | |
| GPT-4o-Video-178KBackbone=Molmo2-8B2026.06 | 56.5 | |
| CapRL-Video-178KBackbone=Molmo2-4B2026.06 | 55.9 | |
| ShareGPT4Video-8B-Video-178KBackbone=Molmo2-8B2026.06 | 55.6 | |
| Tarsier2-7B-Video-178KBackbone=Molmo2-8B2026.06 | 55.3 | |
| Tarsier2-7B-Video-178KBackbone=Molmo2-4B2026.06 | 54.5 | |
| GPT-4o-Video-178KBackbone=Molmo2-4B2026.06 | 53.2 | |
| ShareGPT4Video-8B-Video-178KBackbone=Molmo2-4B2026.06 | 52.7 |