Multimodal Mathematical Reasoning on MathVista (All, GEO, ALG, GPS, TQA)
77.26Accuracy (All)MathVis-Fine
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MathVis-Fine#Params=7B2026.06 | 77.26 | 77.85 | 77.62 | 78.41 | 72.9 | |
| MINT-CoT-7B#Params=7B2026.06 | 73.7 | 74.53 | 73.78 | 75 | 69.35 | |
| MM-Eureka#Params=7B2026.06 | 72.59 | 71.22 | 72.66 | 72.6 | 72.58 | |
| R1-VL-7B#Params=7B2026.06 | 69.63 | 68.87 | 69.66 | 69.71 | 69.35 | |
| InternVL2-8B-MPO#Params=8B2026.06 | 68.52 | 68.87 | 68.91 | 69.71 | 64.52 | |
| Mulberry#Params=7B2026.06 | 68.52 | 67.92 | 68.54 | 68.75 | 67.74 | |
| Claude-3.5 Sonnet2026.06 | 67.41 | 65.09 | 67.79 | 65.38 | 74.19 | |
| LLaVA-OneVision-Qwen2-7b-ov#Params=7B2026.06 | 67.04 | 69.34 | 67.04 | 69.71 | 58.06 | |
| GPT-4o2026.06 | 66.67 | 63.68 | 67.04 | 63.46 | 77.42 | |
| Qwen2.5-VL-7B-Instruct#Params=7B2026.06 | 66.66 | 65.56 | 66.29 | 65.87 | 69.35 | |
| DeepSeek-VL2#Params=4.5B2026.06 | 65.56 | 63.68 | 65.54 | 63.94 | 70.97 | |
| InternVL2-8B#Params=8B2026.06 | 62.59 | 62.26 | 62.92 | 62.5 | 62.9 | |
| Open-R1-Multimodal#Params=7B2026.06 | 54.81 | 52.36 | 54.68 | 53.37 | 59.68 |