Long Video Understanding on MLVU MCQ (test)
81.5AccuracyGemini 2.5 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 2.5 ProCategory=API call only2026.03 | 81.5 | |
| GPT-5Category=API call only2026.03 | 77.7 | |
| Gemini 3 ProCategory=API call only2026.03 | 75.7 | |
| Gemini 2.5 FlashCategory=API call only2026.03 | 75.1 | |
| GPT-5 miniCategory=API call only2026.03 | 69.1 | |
| Claude Sonnet 4.5Category=API call only2026.03 | 64 | |
| Molmo2-4BCategory=Fully open2026.03 | 63 | |
| MolmoPoint-8BCategory=MolmoPoint2026.03 | 62.4 | |
| MiniCPM-V-4.5-8BCategory=Open weights2026.03 | 60.6 | |
| Eagle2.5-8BCategory=Open weights2026.03 | 60.4 | |
| Molmo2-8BCategory=Fully open2026.03 | 60.2 | |
| Qwen3-VL-4BCategory=Open weights2026.03 | 58.4 | |
| Qwen3-VL-8BCategory=Open weights2026.03 | 57.6 | |
| GLM-4.1V-9BCategory=Open weights2026.03 | 56.6 | |
| VideoChat-Flash-7BCategory=Fully open2026.03 | 56 | |
| Molmo2-O-7BCategory=Fully open2026.03 | 55.2 | |
| Keye-VL-1.5-8BCategory=Open weights2026.03 | 53.8 | |
| InternVL3.5-8BCategory=Open weights2026.03 | 53.2 | |
| LLaVA-Video-7BCategory=Fully open2026.03 | 52.8 | |
| PLM-8BCategory=Fully open2026.03 | 52.6 | |
| InternVL3.5-4BCategory=Open weights2026.03 | 52 | |
| PLM-3BCategory=Fully open2026.03 | 48.4 |