Math & Knowledge on MathVista mini
85.9AccuracyQwen3-VL 32B
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-VL 32BModel Source Category=Open-weight VLMs, Model Parameters=32B, Evaluation Mode=Thinking Mode2026.01 | 85.9 | |
| GLM-4.6V-FlashX2026.01 | 82.5 | |
| MMFineReason-4BModel Source Category=Ours, Model Parameters=4B, Evaluation Mode=Thinking Mode2026.01 | 82.2 | |
| Qwen3-VL 30B-A3BModel Source Category=Open-weight VLMs, Model Parameters=30B-A3B, Evaluation Mode=Thinking Mode2026.01 | 81.9 | |
| MMFineReason-8BModel Source Category=Ours, Model Parameters=8B, Evaluation Mode=Thinking Mode2026.01 | 81.7 | |
| Qwen3-VL 8BModel Source Category=Open-weight VLMs, Model Parameters=8B, Evaluation Mode=Thinking Mode2026.01 | 81.4 | |
| OMR 7BModel Source Category=Open-source VLMs, Model Parameters=7B, Evaluation Mode=Thinking Mode2026.01 | 79.5 | |
| Gemini-2.5 FlashModel Source Category=Closed-source VLMs, Evaluation Mode=Thinking Mode2026.01 | 79.4 | |
| GPT5 miniModel Source Category=Closed-source VLMs, Evaluation Mode=Thinking Mode2026.01 | 79.1 | |
| Qwen3-VLParameter Scale=8B2026.01 | 77.2 | |
| Ostrakon-VLParameter Scale=8B2026.01 | 75.4 | |
| InternVL3.5Parameter Scale=8B2026.01 | 75.3 | |
| MMR1 8BModel Source Category=Open-source VLMs, Model Parameters=8B, Evaluation Mode=Thinking Mode2026.01 | 75.3 | |
| MMFineReason-2BModel Source Category=Ours, Model Parameters=2B, Evaluation Mode=Thinking Mode2026.01 | 74.6 | |
| HoneyBee 8BModel Source Category=Open-source VLMs, Model Parameters=8B, Evaluation Mode=Thinking Mode2026.01 | 71.9 | |
| Qwen2.5-VLParameter Scale=72B2026.01 | 68.8 | |
| InternVL2.5-MPOModel Size=8B2025.01 | 68.5 | |
| SAIL-VLModel Size=8B2025.01 | 68.4 | |
| Ovis2.5Parameter Scale=9B2026.01 | 67.7 | |
| SAIL-VLModel Size=2B2025.01 | 62.8 | |
| DeepSeekVL-2Model Size=8B2025.01 | 61.8 | |
| Qwen2-VLModel Size=8B2025.01 | 57.3 | |
| InternVL2.5-MPOModel Size=2B2025.01 | 55.3 | |
| DeepSeekVL-2Model Size=2B2025.01 | 54.5 | |
| Qwen2-VLModel Size=2B2025.01 | 45 |