Visual Math — Overall
Math · 2025-10-20
This MathArena aggregate covers the full Visual Math problem collection. Scores report the percentage of problems answered correctly.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-5.5 | 94.9 |
| GPT-5.4 | 92.5 |
| Kimi K3 | 91.3 |
| Gemini 3.5 Flash | 89.9 |
| Gemini 3.1 Pro Preview | 89.4 |
| Gemini 3.6 Flash | 87.8 |
| GPT-5.2 | 86.5 |
| Gemini 3 Flash Preview | 85.8 |
| Gemini 3 Pro Preview | 84.2 |
| Opus 4.8 | 81.6 |
| Kimi K2.5 | 80.6 |
| GPT-5 | 78.8 |
| GPT-5 Mini | 78.2 |
| Gemini 2.5 Pro | 77.2 |
| GPT-5.1 | 76.9 |
| Sonnet 4.5 | 75.8 |
| Qwen3-VL-235B-A22B-Instruct | 72.5 |
| Opus 4.6 | 72.3 |
| Grok 4 | 70.0 |
| Grok 4.1 Fast | 69.0 |
| GLM-4.5V | 67.6 |
| Grok 4 Fast | 66.8 |