MMMU-Pro w/ python

MMMU-Pro w/ python is a benchmark for understanding images combined with text. The official description says: "MMMU-Pro with python tool, averaged over 4 runs". Below we describe in simple words what benchmarks of this type generally examine.

MMMU-Pro w/ python measures understanding images combined with text. In detail, it checks:

• answering questions about images (figures, photos, charts, memes)
• reading text inside images and connecting it to the question
• visual reasoning such as counting, locating and comparing objects
• combining what is seen with general knowledge

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better.

Official methodology note: MMMU-Pro with python tool, averaged over 4 runs.

What is the MMMU-Pro w/ python benchmark?

MMMU-Pro w/ python is a benchmark for understanding images combined with text. The official description says: 'MMMU-Pro with python tool, averaged over 4 runs'. Below we describe in simple words what benchmarks of this type generally examine.

What does MMMU-Pro w/ python test?

MMMU-Pro w/ python mainly measures understanding images combined with text. Concretely, it checks: answering questions about images (figures, photos, charts, memes); reading text inside images and connecting it to the question; visual reasoning such as counting, locating and comparing objects.

How is the MMMU-Pro w/ python score calculated?

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better. The declared maximum score of this benchmark is 100.0.

Where do I get MMMU-Pro w/ python data and scores?

The official source is https://huggingface.co/moonshotai/Kimi-K3.1-Pro. The official benchmark page/reference is linked on this page. Every score on Open LLM Radar is also available as open data: download the scores JSON (http://open-llm-radar-api:8000/api/v1/benchmarks/mmmu-pro-with-python/scores) or pull it directly from the Open LLM Radar API with any HTTP client.