Multi-Image Understanding on MuirBench 142 (test)
86.1ScoreGemini 3 Pro
Evaluation results
| Method | Links | |
|---|---|---|
| Gemini 3 ProModel Category=API call only2026.01 | 86.1 | |
| GPT-5Model Category=API call only2026.01 | 78.6 | |
| GLM-4.1V-9BModel Category=Open weights only2026.01 | 74.7 | |
| Gemini 2.5 ProModel Category=API call only2026.01 | 74.5 | |
| Gemini 2.5 FlashModel Category=API call only2026.01 | 73.5 | |
| GPT-5 miniModel Category=API call only2026.01 | 71.4 | |
| Qwen3-VL-8BModel Category=Open weights only2026.01 | 64.4 | |
| Qwen3-VL-4BModel Category=Open weights only2026.01 | 63.8 | |
| Molmo2-8BModel Category=Molmo2 family2026.01 | 63.7 | |
| Eagle2.5-8BModel Category=Open weights only2026.01 | 61.8 | |
| Molmo2-4BModel Category=Molmo2 family2026.01 | 60.5 | |
| Claude Sonnet 4.5Model Category=API call only2026.01 | 59.6 | |
| Molmo2-O-7BModel Category=Molmo2 family2026.01 | 58.4 | |
| InternVL3.5-8BModel Category=Open weights only2026.01 | 55.8 | |
| MiniCPM-V-4.5-8BModel Category=Open weights only2026.01 | 53.3 | |
| InternVL3.5-4BModel Category=Open weights only2026.01 | 53.1 | |
| Keye-VL-1.5-8BModel Category=Open weights only2026.01 | 51.2 | |
| PLM-3BModel Category=Open models2026.01 | 25.7 | |
| PLM-8BModel Category=Open models2026.01 | 23.5 |