MMMLU
Multilingual
Knowledge
+
+
+
+
About
MMMLU (Multilingual Massive Multitask Language Understanding) evaluates AI models on general knowledge and reasoning tasks across 57 academic domains translated into 14 languages, testing multilingual generalization beyond English-only performance.
+
+
+
+
Evaluation Stats
Total Models6
Organizations3
Verified Results0
Self-Reported6
+
+
+
+
Benchmark Details
Max Score100
+
+
+
+
Performance Overview
Score distribution and top performers
Score Distribution
6 models
Top Score
91.8%
Average Score
90.3%
High Performers (80%+)
6Top Organizations
#1Google DeepMind
1 model
91.8%
#2Anthropic
4 models
90.2%
#3OpenAI
1 model
89.6%
+
+
+
+
Leaderboard
6 models ranked by performance on MMMLU
| License | Links | ||||
|---|---|---|---|---|---|
| Nov 18, 2025 | Proprietary | 91.8% | |||
| Feb 1, 2026 | Proprietary | 91.1% | |||
| Nov 1, 2025 | Proprietary | 90.8% | |||
| Dec 11, 2025 | Proprietary | 89.6% | |||
| Sep 29, 2025 | Proprietary | 89.5% | |||
| Feb 17, 2026 | Proprietary | 89.3% |
