Multimodal reasoning
MMMU-Pro
A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better.
Rankings
Higher is betterMMMU-Pro — frequently asked questions
- What is MMMU-Pro?
- A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better.
- Which AI model scores highest on MMMU-Pro?
- Gemini 3.5 Flash by Google holds the best MMMU-Pro result among tracked models, at 83.6% (released May 19 2026). Higher scores are better on this benchmark.
- What are the top 5 models on MMMU-Pro?
- 1. Gemini 3.5 Flash (Google) — 83.6%; 2. Kimi K3 (Moonshot AI) — 81.6%; 3. Gemini 3.0 Flash (Google) — 81.2%; 3. GPT-5.5 (OpenAI) — 81.2%; 5. Gemini 3.1 Pro (Google) — 80.5%.
- How many models have a published MMMU-Pro score?
- 9 tracked models have a published MMMU-Pro score. Scores are the figures reported by each lab at that model's release, so this page is a record of results over time rather than a re-run leaderboard.
- What is the best open model on MMMU-Pro?
- Qwen3.5 by Qwen is the highest-ranked model with downloadable weights on MMMU-Pro, scoring 79% at rank 6 overall.