// BENCHMARK

MMMU-Pro (Vision) benchmark

AI model leaderboard for the MMMU-Pro (Vision) benchmark. Compare how large language models score on MMMU-Pro (Vision), see the full ranking, and understand what this AI benchmark measures. Magistral Medium 1.1 currently leads with 52.1. Vision split of MMMU-Pro — questions are rendered as screenshots so the model must read the image text.

Leaderboard

# Model Organization Score Variant Source
#1 Magistral Medium 1.1 Mistral AI 52.1 official ↗
#2 Mistral Small 3.1 Mistral AI 42.4 cited-magistral official ↗
#3 Mistral Medium 3 Mistral AI 39.7 cited-magistral official ↗
#4 Magistral Small 1.1 Mistral AI 39.2 official ↗

Frequently asked questions about MMMU-Pro (Vision)

What is the MMMU-Pro (Vision) benchmark?

Vision split of MMMU-Pro — questions are rendered as screenshots so the model must read the image text.

How is the MMMU-Pro (Vision) benchmark scored?

MMMU-Pro (Vision) is scored using the accuracy metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.

Which AI model scores highest on MMMU-Pro (Vision)?

As of the latest reported scores on GenAIList, Magistral Medium 1.1 achieves the highest result on MMMU-Pro (Vision) with a score of 52.1.

Is a higher MMMU-Pro (Vision) score better?

Yes. On MMMU-Pro (Vision) a higher score indicates better performance, so models near the top of the leaderboard are the strongest.