Evals for Every Language - Language su: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 70.66 |
| 2 | Claude Opus 4.8 | 69.97 |
| 3 | Claude Sonnet 4.6 | 69.64 |
| 4 | Gemini 3.1 Pro (Preview) | 69.43 |
| 5 | Claude Sonnet 4.5 | 68.89 |
| 6 | Qwen 3.6 Plus | 68.18 |
| 7 | Grok 4.20 | 67.61 |
| 8 | GPT-5.3 | 67.4 |
| 9 | Claude Sonnet 4 | 67.34 |
| 10 | Gemini 3.1 Flash Lite | 66.97 |
| 11 | GPT-5.2 | 66.57 |
| 12 | GPT-5.5 | 66.24 |
| 13 | GPT-5.1 | 66.2 |
| 14 | Gemini 2.5 Pro | 66.11 |
| 15 | DeepSeek V3.1 | 65.93 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-su · How It Works · Data refreshed daily, snapshot 2026-09-05.