MMMLU (OpenAI Multilingual MMLU) — OpenAI's human-translated multilingual MMLU across 14 languages.
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Claude Mythos 5 | Anthropic | 92.7 | 1 | 2026-08-23 |
| 2 | Gemini 3.1 Pro | 92.6 | 4 | 2026-08-23 | |
| 3 | Gemini 3 Flash Preview | 91.8 | 1 | 2026-08-23 | |
| 4 | Gemini 3 Pro | 91.8 | 3 | 2026-08-23 | |
| 5 | Claude Opus 4.7 | Anthropic | 91.5 | 3 | 2026-08-23 |
| 6 | Claude Opus 4.6 | Anthropic | 91.1 | 4 | 2026-08-23 |
| 7 | GPT-5.1 | OpenAI | 91 | 1 | 2026-07-29 |
| 8 | Claude Opus 4.5 | Anthropic | 90.8 | 3 | 2026-08-23 |
| 9 | Qwen3.7 Max | Alibaba | 90.3 | 1 | 2026-08-23 |
| 10 | GPT-5.2 | OpenAI | 89.6 | 2 | 2026-08-23 |
| 11 | Claude Opus 4.1 | Anthropic | 89.5 | 2 | 2026-08-23 |
| 12 | Qwen3.6 Plus | Alibaba | 89.5 | 1 | 2026-08-23 |
| 13 | Claude Sonnet 4.6 | Anthropic | 89.3 | 2 | 2026-08-23 |
| 14 | Claude Sonnet 4.5 | Anthropic | 89.1 | 2 | 2026-08-23 |
| 15 | Qwen3.7 Plus Preview | Alibaba | 89 | 1 | 2026-08-23 |
| 16 | Gemini 3.1 Flash Lite Preview | 88.9 | 1 | 2026-08-23 | |
| 17 | Claude 4 Opus | Anthropic | 88.8 | 2 | 2026-08-23 |
| 18 | Qwen3.5 397B A17B | Alibaba | 88.5 | 1 | 2026-08-23 |
| 19 | Gemma 4 31B | 88.4 | 27 | 2026-08-23 | |
| 20 | o1 | OpenAI | 87.7 | 1 | 2026-08-23 |
| 21 | GPT-4.1 | OpenAI | 87.3 | 1 | 2026-08-23 |
| 22 | Qwen3.5 122B A10B | Alibaba | 86.7 | 1 | 2026-08-23 |
| 23 | Qwen3 235B A22B | Alibaba | 86.7 | 1 | 2026-08-23 |
| 24 | Claude Sonnet 4 | Anthropic | 86.5 | 2 | 2026-08-23 |
| 25 | Gemma 4 26B A4B | 86.3 | 27 | 2026-08-23 | |
| 26 | Claude 3.7 Sonnet | Anthropic | 86.1 | 1 | 2026-08-23 |
| 27 | Qwen3.5 27B | Alibaba | 85.9 | 1 | 2026-08-23 |
| 28 | Mistral Large 3 675B Base 2512 | Mistral | 85.5 | 1 | 2026-08-23 |
| 29 | Mistral Large | Mistral | 85.5 | 1 | 2026-08-23 |
| 30 | Qwen3.5 35B A3B | Alibaba | 85.2 | 1 | 2026-08-23 |
| 31 | GPT-4.5 Preview | OpenAI | 85.1 | 1 | 2026-08-23 |
| 32 | Gemma 4 12B | 83.4 | 27 | 2026-08-23 | |
| 33 | Claude Haiku 4.5 | Anthropic | 83 | 1 | 2026-08-23 |
| 34 | GPT-4o | OpenAI | 81.4 | 1 | 2026-08-23 |
| 35 | Qwen3.5 9B | Alibaba | 81.2 | 1 | 2026-08-23 |
| 36 | Hy3 preview-Base | Tencent | 80.2 | 2 | 2026-06-05 |
| 37 | DeepSeek-V3-Base | DeepSeek | 79.5 | 2 | 2026-06-05 |
| 38 | DeepSeek-V3 | DeepSeek | 79.4 | 1 | 2026-07-06 |
| 39 | GLM-4.5-Base | Z.ai | 79.3 | 2 | 2026-06-05 |
| 40 | Qwen3 32B | Alibaba | 79 | 1 | 2026-06-15 |
| 41 | Qwen3 30B A3B | Alibaba | 79 | 1 | 2026-06-15 |
| 42 | GPT-4.1 mini | OpenAI | 78.5 | 1 | 2026-08-23 |
| 43 | Seed Oss 36B Instruct | ByteDance | 78.4 | 1 | 2026-06-15 |
| 44 | Kimi K2 Base | Moonshot | 77.6 | 2 | 2026-06-05 |
| 45 | Gemma 4 E4B | 76.6 | 32 | 2026-08-23 | |
| 46 | Qwen3.5 4B | Alibaba | 76.1 | 1 | 2026-08-23 |
| 47 | Qwen2.5 Instruct 72B | Alibaba | 74.8 | 1 | 2026-07-06 |
| 48 | Mistral Large 3 | Mistral | 74.2 | 1 | 2026-08-23 |
| 49 | Llama 3.1 Instruct 405B | Meta | 73.8 | 1 | 2026-07-06 |
| 50 | Granite 4.1 30B | IBM | 73.7 | 1 | 2026-06-15 |
| 51 | Granite 4.1 30B Base | IBM | 73.4 | 2 | 2026-06-15 |
| 52 | Gemma 3 27B | 70.7 | 27 | 2026-07-10 | |
| 53 | Phi 3.5 MoE Instruct | Microsoft | 69.9 | 1 | 2026-08-23 |
| 54 | Gemma 4 E2B | 67.4 | 37 | 2026-08-23 | |
| 55 | GPT-4.1 nano | OpenAI | 66.9 | 1 | 2026-08-23 |
| 56 | Qwen2.5 7B Base | Alibaba | 65.5 | 1 | 2026-05-03 |
| 57 | Granite 4.1 8B | IBM | 64.8 | 1 | 2026-06-15 |
| 58 | Granite 4.1 8B Base | IBM | 64.7 | 2 | 2026-06-15 |
| 59 | DeepSeek-V2 | DeepSeek | 64 | 1 | 2026-07-06 |
| 60 | Qwen3.5 2B | Alibaba | 63.1 | 1 | 2026-08-23 |
| 61 | Qwen2.5 3B Base | Alibaba | 59.6 | 1 | 2026-05-03 |
| 62 | Granite 4.1 3B Base | IBM | 56.6 | 1 | 2026-06-15 |
| 63 | Phi 3.5 Mini Instruct | Microsoft | 55.4 | 1 | 2026-08-23 |
| 64 | LFM2 2.6B | Liquid AI | 55.4 | 1 | 2026-08-09 |
| 65 | Qwen3 4B Base Dense | Alibaba | 50.2 | 1 | 2026-05-03 |
| 66 | Gemma 3 4B | 50.1 | 1 | 2026-08-09 | |
| 67 | Llama 3.2 3B Instruct | Meta | 47.9 | 1 | 2026-08-09 |
| 68 | LFM2 1.2B | Liquid AI | 46.7 | 1 | 2026-08-09 |
| 69 | Qwen3 1.7B | Alibaba | 46.5 | 1 | 2026-08-09 |
| 70 | Qwen3.5 0.8B | Alibaba | 44.3 | 1 | 2026-08-23 |
| 71 | LFM2-700M | Liquid AI | 43.3 | 1 | 2026-08-09 |
| 72 | LFM2-350M | Liquid AI | 38 | 1 | 2026-08-09 |
| 73 | Gemma 3 1B IT | 34.4 | 1 | 2026-08-09 | |
| 74 | Qwen3 0.6B | Alibaba | 30.8 | 1 | 2026-08-09 |
| 75 | Gemma 3 1B Base Dense | 26.6 | 1 | 2026-05-03 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.