Evals for Every Language - MMLU: leaderboard
LanguageBench MMLU accuracy averaged across available languages and scaled to 0-100. Translation uses spBLEU only, without mixing chrF; higher is better.
Metric: Average Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 69 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 98.69 | #58 |
| 2 | Claude Sonnet 4.6 | 98.47 | #90 |
| 3 | Claude Opus 4.8 | 98.31 | #37 |
| 4 | GPT-5.5 | 98.21 | #31 |
| 5 | DeepSeek V3.1 | 97.67 | #264 |
| 6 | Grok 4.20 | 96.65 | #175 |
| 7 | Gemini 3.1 Flash Lite | 96.38 | #191 |
| 8 | Claude Opus 4.7 | 95.33 | #49 |
| 9 | GPT-5.3 | 95.18 | #103 |
| 10 | GPT-5.2 | 95.15 | #110 |
| 11 | Qwen 3.6 Plus | 94.86 | #120 |
| 12 | GPT-5 | 94.83 | #98 |
| 13 | GPT-5.1 | 93.32 | #135 |
| 14 | GPT-5.4 | 93.09 | #79 |
| 15 | Claude Sonnet 4 | 90.23 | #199 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-mmlu · How It Works · Data refreshed daily, snapshot 2026-10-20.