Which AI model scores highest on MMMLU?

Gemini 3.1 Pro by Google holds the best MMMLU result among tracked models, at 92.6% (released Feb 19 2026). Higher scores are better on this benchmark.

What are the top 5 models on MMMLU?

1. Gemini 3.1 Pro (Google) — 92.6%; 2. Gemini 3.0 Pro (Google) — 91.8%; 2. Gemini 3.0 Flash (Google) — 91.8%; 4. GPT-5.2 (OpenAI) — 89.6%; 5. Gemini 3.1 Flash-Lite (Google) — 88.9%.

How many models have a published MMMLU score?

12 tracked models have a published MMMLU score. Scores are collected from published lab reports and benchmark sources, with recorded sources credited alongside the results. We do not re-run the evaluations.

What is the best open model on MMMLU?

Gemma 4 by Google is the highest-ranked model with downloadable weights on MMMLU, scoring 88.4% at rank 6 overall.

Knowledge

MMMLU

average accuracy on multilingual translations of MMLU

Rankings

Higher is better

Gemini 3.1 Flash-Lite

MMMLU — frequently asked questions

: average accuracy on multilingual translations of MMLU
: Gemini 3.1 Pro by Google holds the best MMMLU result among tracked models, at 92.6% (released Feb 19 2026). Higher scores are better on this benchmark.
: 1. Gemini 3.1 Pro (Google) — 92.6%; 2. Gemini 3.0 Pro (Google) — 91.8%; 2. Gemini 3.0 Flash (Google) — 91.8%; 4. GPT-5.2 (OpenAI) — 89.6%; 5. Gemini 3.1 Flash-Lite (Google) — 88.9%.
: 12 tracked models have a published MMMLU score. Scores are collected from published lab reports and benchmark sources, with recorded sources credited alongside the results. We do not re-run the evaluations.
: Gemma 4 by Google is the highest-ranked model with downloadable weights on MMMLU, scoring 88.4% at rank 6 overall.