Evals for Every Language - Language pag: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 72.21 |
| 2 | GPT-5.5 | 67.84 |
| 3 | Claude Opus 4.7 | 66.97 |
| 4 | Gemini 2.5 Pro | 66.8 |
| 5 | Claude Sonnet 4.6 | 64.27 |
| 6 | Gemini 3.1 Flash Lite | 64.19 |
| 7 | GPT-5.4 | 63.9 |
| 8 | Claude Opus 4.8 | 63.49 |
| 9 | Qwen 3.6 Plus | 59.39 |
| 10 | GPT-5.3 | 58.45 |
| 11 | Grok 4.20 | 57.62 |
| 12 | Claude Sonnet 4.5 | 57.52 |
| 13 | DeepSeek V3.2 Exp | 55.46 |
| 14 | DeepSeek V3.1 | 55.22 |
| 15 | GPT-5.2 | 55.09 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-pag · How It Works · Data refreshed daily, snapshot 2026-09-05.