Video Understanding on MLVU MCQ (test)
81.5AccuracyGemini 2.5 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 2.5 ProAccess Type=API call only2026.01 | 81.5 | |
| GPT-5Access Type=API call only2026.01 | 77.7 | |
| Gemini 3 ProAccess Type=API call only2026.01 | 75.7 | |
| Gemini 2.5 FlashAccess Type=API call only2026.01 | 75.1 | |
| GPT-5 miniAccess Type=API call only2026.01 | 69.1 | |
| Claude Sonnet 4.5Access Type=API call only2026.01 | 64 | |
| Molmo2-4BAccess Type=Molmo2 family2026.01 | 63 | |
| MiniCPM-V-4.5-8BAccess Type=Open weights only2026.01 | 60.6 | |
| Eagle2.5-8BAccess Type=Open weights only2026.01 | 60.4 | |
| Molmo2-8BAccess Type=Molmo2 family2026.01 | 60.2 | |
| Qwen3-VL-4BAccess Type=Open weights only2026.01 | 58.4 | |
| Qwen3-VL-8BAccess Type=Open weights only2026.01 | 57.6 | |
| GLM-4.1V-9BAccess Type=Open weights only2026.01 | 56.6 | |
| VideoChat-Flash-7BAccess Type=Open models2026.01 | 56 | |
| Molmo2-O-7BAccess Type=Molmo2 family2026.01 | 55.2 | |
| Keye-VL-1.5-8BAccess Type=Open weights only2026.01 | 53.8 | |
| InternVL3.5-8BAccess Type=Open weights only2026.01 | 53.2 | |
| LLaVA-Video-7BAccess Type=Open models2026.01 | 52.8 | |
| PLM-8BAccess Type=Open models2026.01 | 52.6 | |
| InternVL3.5-4BAccess Type=Open weights only2026.01 | 52 | |
| PLM-3BAccess Type=Open models2026.01 | 48.4 |