Skip to content

MMMU-Pro with Python

Tool-augmented MMMU-Pro variant that allows Python assistance during multimodal reasoning.

9 models

Higher is better.

  1. 1GPT-5.6 SolOpenAI84.6
  2. 2Kimi K3Moonshot AI83.4
  3. 3GPT-5.5OpenAI83.2
  4. 4GPT-5.4OpenAI82.1
  5. 5GPT-5.6 TerraOpenAI82.0
  6. 6Kimi K2.6Moonshot AI80.1
  7. 7GPT-5.6 LunaOpenAI79.5
  8. 8GPT-5.4 miniOpenAI78.0
  9. 9GPT-5.4 nanoOpenAI69.5
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Image + text question answering with Python
Tasks
Multimodal academic reasoning
Authors
OpenAI (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.