Multimodal Mathematical Reasoning on MathVerse mini (test)
70.94T-Dominant ScoreClaude-4-Sonnet
Evaluation Results
| Method | Links | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Claude-4-SonnetModel Category=Proprietary2026.02 | 70.94 | 49.85 | 65.99 | 56.98 | 50.38 | 52.54 | 18.53 | — | — | — | — | — | — | — | — | |
| Gemini-2.5-ProModel Category=Proprietary2026.02 | 65.23 | 50.76 | 56.98 | 55.33 | 50.76 | 50.76 | 31.73 | — | — | — | — | — | — | — | — | |
| Guided-GRPOBackbone=Qwen3-8B, Training=G-GRPO+SFT2026.02 | 61.68 | 51.07 | 53.43 | 53.93 | 48.98 | 48.73 | 42.01 | — | — | — | — | — | — | — | — | |
| GPT-4oModel Category=Proprietary2026.02 | 61.17 | 48.88 | 54.7 | 52.92 | 46.45 | 46.95 | 36.93 | — | — | — | — | — | — | — | — | |
| Base (Qwen3-8B)Backbone=Qwen3-8B2026.02 | 58.38 | 46.83 | 47.08 | 48.6 | 45.05 | 44.42 | 37.69 | — | — | — | — | — | — | — | — | |
| Vision-R1-32BModel Category=Open-Source2026.02 | 57.74 | 50.66 | 49.87 | 53.05 | 50 | 48.6 | 43.91 | — | — | — | — | — | — | — | — | |
| MMR1-32BModel Category=Open-Source2026.02 | 54.95 | 46.35 | 47.34 | 46.95 | 45.18 | 44.04 | 40.61 | — | — | — | — | — | — | — | — | |
| GPT-4V2026.01 | 54.7 | — | 48.7 | 41.4 | 34.9 | 34.4 | 31.6 | 39.4 | 54.4 | 63.1 | 56.6 | 60.3 | 51.4 | 50.8 | 50.3 | |
| VL-Rethinker-7BModel Category=Open-Source2026.02 | 54.44 | 47.66 | 48.35 | 50.38 | 45.94 | 45.56 | 42.01 | — | — | — | — | — | — | — | — | |
| MathFlow GPT-4V2026.01 | 51.1 | — | 48.5 | 46.4 | 40.3 | 37.4 | 39 | 43.8 | 56.7 | 65.2 | 58.9 | 62.1 | 53.7 | 52.1 | 52.5 | |
| Qwen-VL-MaxModel Category=Proprietary2026.02 | 48.98 | 35.91 | 28.93 | 39.47 | 30.08 | 30.08 | 30.94 | — | — | — | — | — | — | — | — | |
| Math-LLaVA-13BModel Category=Open-Source2026.02 | 48.98 | 33.68 | 31.47 | 39.85 | 28.05 | 27.66 | 23.86 | — | — | — | — | — | — | — | — | |
| LLaVA-v1.5-13BModel Category=Open-Source2026.02 | 48.22 | 35.48 | 34.77 | 39.47 | 36.04 | 34.14 | 19.54 | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-32BModel Category=Open-Source2026.02 | 44.29 | 38.88 | 45.18 | 39.97 | 39.09 | 40.1 | 30.96 | — | — | — | — | — | — | — | — | |
| InternVL2.5-26BModel Category=Open-Source2026.02 | 43.9 | 39.34 | 41.5 | 38.71 | 41.5 | 41.37 | 31.22 | — | — | — | — | — | — | — | — | |
| COGFLOW-7BParameters=7B2026.01 | 41.9 | — | 40.1 | 37 | 42.4 | 44.8 | 26.3 | 39.5 | 53.9 | 60.7 | 51.2 | 52.3 | 55 | 58.7 | 44.2 | |
| SVE-Math-Qwen2.5-7BBackbone=Qwen2.5, Parameters=7B2026.01 | 37.6 | — | — | 36.8 | 34.9 | 31.5 | 16 | 31.4 | — | — | — | — | — | — | — | |
| MultiMath-7BParameters=7B2026.01 | 34.8 | — | — | 30.8 | 28.1 | 25.9 | 15 | 26.9 | — | — | — | — | — | — | — | |
| Qwen-VL-Max2026.01 | 30.7 | — | 28.9 | 26.1 | 24.1 | 24.1 | 21.4 | 25.3 | 37.2 | 42.8 | 37.7 | 47.9 | 33.6 | 35.9 | 35.9 | |
| Llama-3.2-11BModel Category=Open-Source2026.02 | 27.16 | 24.29 | 25.76 | 24.87 | 24.87 | 24.49 | 20.05 | — | — | — | — | — | — | — | — | |
| Gemini-Pro2026.01 | 26.3 | — | 27.3 | 23.5 | 23 | 22.3 | 22.2 | 23.5 | 35.3 | 39.8 | 34.7 | 44.5 | 32 | 36.8 | 33.3 | |
| Phi-3-vision-128kModel Category=Open-Source2026.02 | 22.97 | 17.46 | 20.05 | 17.64 | 17.26 | 15.86 | 13.58 | — | — | — | — | — | — | — | — | |
| Math-LLaVA-13BParameters=13B2026.01 | 22.8 | — | — | 21.8 | 21.1 | 19.2 | 15.4 | 20.1 | — | — | — | — | — | — | — | |
| InternLM-XC2-7BParameters=7B2026.01 | 22.3 | — | 16.5 | 17 | 15.7 | 16.4 | 11 | 16.5 | 25.9 | 36.9 | 28.3 | 42.5 | 20.1 | 24.4 | 19.8 | |
| SPHINX-MoE-56BParameters=56B2026.01 | 22.2 | — | 18.3 | 16.4 | 14.8 | 12.6 | 9.1 | 15.6 | 25.8 | 33.3 | 21.9 | 40.7 | 21.1 | 19.6 | 18.3 | |
| Qwen-VL-Plus2026.01 | 15.7 | — | 14.5 | 11.1 | 9 | 13 | 10 | 11.8 | 21.3 | 26 | 21.2 | 25.2 | 18.5 | 19.1 | 21.8 | |
| DVLR-14BParameters=14B2026.01 | — | — | — | — | — | — | — | — | 48.1 | 54.3 | 49 | — | 46.3 | 47.2 | 43.8 | |
| GRPOModel=Qwen2.5-VL-3B, Paradigm=GRPO2026.02 | — | — | — | — | — | — | — | 48.67 | — | — | — | — | — | — | — | |
| GRPOModel=Qwen2.5-VL-7B, Paradigm=GRPO2026.02 | — | — | — | — | — | — | — | 57.16 | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B2026.01 | — | — | — | — | — | — | — | 44.8 | — | — | — | — | — | — | — | |
| SFTModel=Qwen2.5-VL-3B, Paradigm=SFT2026.02 | — | — | — | — | — | — | — | 49.54 | — | — | — | — | — | — | — | |
| SFTModel=Qwen2.5-VL-7B, Paradigm=SFT2026.02 | — | — | — | — | — | — | — | 58.39 | — | — | — | — | — | — | — | |
| SFT-MModel=Qwen2.5-VL-3B, Paradigm=SFT-M2026.02 | — | — | — | — | — | — | — | 51.51 | — | — | — | — | — | — | — | |
| SFT-MModel=Qwen2.5-VL-7B, Paradigm=SFT-M2026.02 | — | — | — | — | — | — | — | 59.31 | — | — | — | — | — | — | — | |
| SFT-RSModel=Qwen2.5-VL-3B, Paradigm=SFT-RS2026.02 | — | — | — | — | — | — | — | 48.72 | — | — | — | — | — | — | — | |
| SFT-RSModel=Qwen2.5-VL-7B, Paradigm=SFT-RS2026.02 | — | — | — | — | — | — | — | 55.87 | — | — | — | — | — | — | — | |
| SophiaVL-R1-7BParameters=7B2026.01 | — | — | — | — | — | — | — | — | 48.8 | 45.4 | 43.9 | — | 45.1 | 58.5 | 51.3 | |
| VideoAuto-R1Backbone=Qwen2.5-VL-7B2026.01 | — | — | — | — | — | — | — | 46.9 | — | — | — | — | — | — | — |