Benchmark

MMMU-Pro

Measures: College-level multimodal multiple-choice questions (up to 10 options), filtered so text-only models cannot answer them, plus a vision-only setting where the whole question sits inside a screenshot or photo; scored by accuracy.

Maintainer
MMMU team (Xiang Yue et al.)
Introduced
2024
Status
Active
Saturation
Near-ceiling
Typical provenance
Mixed
In scoring set
Yes

PaperOfficial leaderboard

Leaderboard

Independent and third-party results only.

No independent or third-party results yet. Show all results

MMMU-Pro: leaderboard and what it measures · supermarket.dev