MathVision with Python
A tool-augmented MathVision variant that permits Python during visual mathematics reasoning.
5 models
Higher is better.
- 1Kimi K3Moonshot AI97.8
- 2Qwen3.8 MaxAlibaba97.7
- 3Qwen3.8-Omni-FlashAlibaba96.2
- 4Qwen3.8-Flash-NextAlibaba95.7
- 5Qwen3.8-27BAlibaba94.6
About this benchmark
- Metric
- Image and mathematics reasoning with tools
- Tasks
- Visual mathematics problems with Python
- Authors
- Moonshot AI / MathVision authors (2026)
Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.