Evals for Every Language - MMLU — leaderboard
Metric: Average Score (%). Source: huggingface.co. 69 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 98.69 |
| 2 | Claude Sonnet 4.6 | 98.47 |
| 3 | Claude Opus 4.8 | 98.31 |
| 4 | GPT-5.5 | 98.21 |
| 5 | DeepSeek V3.1 | 97.67 |
| 6 | Grok 4.20 | 96.65 |
| 7 | Gemini 3.1 Flash Lite | 96.38 |
| 8 | Claude Opus 4.7 | 95.33 |
| 9 | nova-2-lite-v1 | 95.33 |
| 10 | GPT-5.2 | 95.15 |
| 11 | Qwen 3.6 Plus | 94.86 |
| 12 | GPT-5 | 94.83 |
| 13 | GPT-5.1 | 93.32 |
| 14 | GPT-5.4 | 93.09 |
| 15 | Claude Sonnet 4 | 90.23 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-mmlu · How the rankings work · Data refreshed daily, snapshot 2026-07-22.