language benchmark
MMMLU Leaderboard
MMMLU is a multilingual dataset released by OpenAI with professionally translated MMLU test questions across 14 languages and 57 subjects.
Updated Sep 7, 2026
Models49
Model coverage49
MetricScore
EvidenceB
MMMLU Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAN | Score92.70% | Percentile100.00% | Participants49 | EvidenceC | Evaluated |
| Rank02 | ModelGO | Score92.60% | Percentile97.92% | Participants49 | EvidenceC | Evaluated |
| Rank03 | ModelGO | Score91.80% | Percentile95.83% | Participants49 | EvidenceC | Evaluated |
| Rank04 | ModelGO | Score91.80% | Percentile93.75% | Participants49 | EvidenceC | Evaluated |
| Rank05 | ModelAN | Score91.50% | Percentile91.67% | Participants49 | EvidenceC | Evaluated |
| Rank06 | ModelAN | Score91.10% | Percentile89.58% | Participants49 | EvidenceC | Evaluated |
| Rank07 | ModelAN | Score90.80% | Percentile87.50% | Participants49 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score90.30% | Percentile85.42% | Participants49 | EvidenceC | Evaluated |
| Rank09 | ModelOP | Score89.60% | Percentile83.33% | Participants49 | EvidenceC | Evaluated |
| Rank10 | ModelAN | Score89.50% | Percentile81.25% | Participants49 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score89.50% | Percentile79.17% | Participants49 | EvidenceC | Evaluated |
| Rank12 | ModelAN | Score89.30% | Percentile77.08% | Participants49 | EvidenceC | Evaluated |
| Rank13 | ModelAN | Score89.10% | Percentile75.00% | Participants49 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score89.00% | Percentile72.92% | Participants49 | EvidenceC | Evaluated |
| Rank15 | ModelGO | Score88.90% | Percentile70.83% | Participants49 | EvidenceC | Evaluated |
| Rank16 | ModelAN | Score88.80% | Percentile68.75% | Participants49 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score88.50% | Percentile66.67% | Participants49 | EvidenceC | Evaluated |
| Rank18 | ModelGO | Score88.40% | Percentile64.58% | Participants49 | EvidenceC | Evaluated |
| Rank19 | ModelOP | Score87.70% | Percentile62.50% | Participants49 | EvidenceC | Evaluated |
| Rank20 | ModelOP | Score87.30% | Percentile60.42% | Participants49 | EvidenceC | Evaluated |
| Rank21 | ModelAC | Score86.70% | Percentile58.33% | Participants49 | EvidenceC | Evaluated |
| Rank22 | ModelAC | Score86.70% | Percentile56.25% | Participants49 | EvidenceC | Evaluated |
| Rank23 | ModelAN | Score86.50% | Percentile54.17% | Participants49 | EvidenceC | Evaluated |
| Rank24 | ModelGO | Score86.30% | Percentile52.08% | Participants49 | EvidenceC | Evaluated |
| Rank25 | ModelAN | Score86.10% | Percentile50.00% | Participants49 | EvidenceC | Evaluated |
| Rank26 | ModelAC | Score85.90% | Percentile47.92% | Participants49 | EvidenceC | Evaluated |
| Rank27 | ModelLA | Score85.70% | Percentile45.83% | Participants49 | EvidenceC | Evaluated |
| Rank28 | ModelMA | Score85.50% | Percentile43.75% | Participants49 | EvidenceC | Evaluated |
| Rank29 | ModelMA | Score85.50% | Percentile41.67% | Participants49 | EvidenceC | Evaluated |
| Rank30 | ModelMA | Score85.50% | Percentile39.58% | Participants49 | EvidenceC | Evaluated |
MMMLU Highlights
The leading models and scores on this benchmark.
MMMLU Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MMMLU
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MMMLU?
What MMMLU measures and how its scores work.
MMMLU is a language benchmark containing approximately 15,908 multiple-choice questions per language in Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese.
It measures Score on multiple-choice questions, reported as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMMLU
- Modality
- text
- Primary category
- language
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MMMLU.
Which model scores highest on MMMLU?
Claude Mythos Preview is currently ranked first with 92.70%.
What are the top three models on MMMLU?
The current leaders are Claude Mythos Preview (92.70%), Gemini 3.1 Pro (92.60%), and Gemini 3 Pro (91.80%).
Which MMMLU model has the lowest official input price?
GPT-4.1 nano has the lowest matched official input price at $0.10 input / $0.40 output per 1M tokens.
Which models are fastest among MMMLU results?
The fastest matched records are GPT-4.1 mini (467.39 tok/s via OpenAI), GPT-4.1 nano (138.14 tok/s via OpenAI), and GPT-4o (132.00 tok/s via OpenAI).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.