Evals for Every Language - Language st: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 70.62 |
| 2 | Claude Opus 4.7 | 68.11 |
| 3 | Gemini 2.5 Pro | 67.85 |
| 4 | GPT-5.3 | 67.57 |
| 5 | GPT-5.4 | 67.54 |
| 6 | Gemini 3.1 Flash Lite | 67.38 |
| 7 | GPT-5.5 | 66.48 |
| 8 | Claude Sonnet 4.6 | 66.28 |
| 9 | GPT-5 | 65.17 |
| 10 | Claude Sonnet 4.5 | 64.15 |
| 11 | Qwen 3.6 Plus | 64.11 |
| 12 | GPT-5.1 | 63.91 |
| 13 | Claude Opus 4.8 | 63.73 |
| 14 | GPT-5.2 | 62.99 |
| 15 | Grok 4.20 | 62.87 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-st · How It Works · Data refreshed daily, snapshot 2026-09-05.