Video Understanding on MotionBench (val)
65.4AccuracyGPT-5
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5Access Type=API call only2026.01 | 65.4 | |
| GPT-5Category=API call only2026.03 | 65.4 | |
| Gemini 3 ProAccess Type=API call only2026.01 | 62.6 | |
| Gemini 3 ProCategory=API call only2026.03 | 62.6 | |
| Molmo2-8BAccess Type=Molmo2 family2026.01 | 62.2 | |
| Molmo2-8BCategory=Fully open2026.03 | 62.2 | |
| Gemini 2.5 ProAccess Type=API call only2026.01 | 62 | |
| Gemini 2.5 ProCategory=API call only2026.03 | 62 | |
| Molmo2-4BAccess Type=Molmo2 family2026.01 | 61.6 | |
| Molmo2-4BCategory=Fully open2026.03 | 61.6 | |
| PLM-8BAccess Type=Open models2026.01 | 61.4 | |
| PLM-8BModel=PLM-8B, Pruning Strength=None2026.03 | 61.4 | |
| PLM-8BCategory=Fully open2026.03 | 61.4 | |
| MolmoPoint-8BCategory=MolmoPoint2026.03 | 61.4 | |
| BaselineModel=Baseline, Pruning Strength=0%2026.03 | 61 | |
| VideoChat-Flash-7BAccess Type=Open models2026.01 | 60.6 | |
| Molmo2-O-7BAccess Type=Molmo2 family2026.01 | 60.6 | |
| VideoChat-Flash-7BCategory=Fully open2026.03 | 60.6 | |
| Molmo2-O-7BCategory=Fully open2026.03 | 60.6 | |
| PLM-3BAccess Type=Open models2026.01 | 60.4 | |
| PLM-3BCategory=Fully open2026.03 | 60.4 | |
| GPT-5 miniAccess Type=API call only2026.01 | 59.9 | |
| GPT-5 miniCategory=API call only2026.03 | 59.9 | |
| MiniCPM-V-4.5-8BAccess Type=Open weights only2026.01 | 59.7 | |
| MiniCPM-V-4.5-8BCategory=Open weights2026.03 | 59.7 | |
| Gemini 2.5 FlashAccess Type=API call only2026.01 | 59.3 | |
| Gemini 2.5 FlashCategory=API call only2026.03 | 59.3 | |
| STTSModel=STTS, Pruning Strength=30%2026.03 | 59.2 | |
| STTSModel=STTS, Pruning Strength=40%2026.03 | 59.2 | |
| GLM-4.1V-9BAccess Type=Open weights only2026.01 | 59 | |
| GLM-4.1V-9BCategory=Open weights2026.03 | 59 | |
| Qwen3-VL-4BAccess Type=Open weights only2026.01 | 58.6 | |
| Qwen3-VL-4BModel=Qwen3-VL-4B, Pruning Strength=None2026.03 | 58.6 | |
| Qwen3-VL-4BCategory=Open weights2026.03 | 58.6 | |
| Claude Sonnet 4.5Access Type=API call only2026.01 | 58.5 | |
| Claude Sonnet 4.5Category=API call only2026.03 | 58.5 | |
| STTSModel=STTS, Pruning Strength=50%2026.03 | 58.2 | |
| Qwen3-VL-8BAccess Type=Open weights only2026.01 | 56.9 | |
| Qwen3-VL-8BCategory=Open weights2026.03 | 56.9 | |
| InternVL3.5-8BAccess Type=Open weights only2026.01 | 56.6 | |
| InternVL3.5-8BModel=InternVL3.5-8B, Pruning Strength=None2026.03 | 56.6 | |
| InternVL3.5-8BCategory=Open weights2026.03 | 56.6 | |
| InternVL3.5-4BAccess Type=Open weights only2026.01 | 56.5 | |
| InternVL3.5-4BCategory=Open weights2026.03 | 56.5 | |
| Eagle2.5-8BAccess Type=Open weights only2026.01 | 55.7 | |
| Eagle2.5-8BCategory=Open weights2026.03 | 55.7 | |
| Keye-VL-1.5-8BAccess Type=Open weights only2026.01 | 55.1 | |
| Keye-VL-1.5-8BCategory=Open weights2026.03 | 55.1 | |
| LLaVA-Video-7BAccess Type=Open models2026.01 | 54.2 | |
| LLaVA-Video-7BCategory=Fully open2026.03 | 54.2 |