Evals for Every Language - Language af: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 79.41 |
| 2 | Claude Sonnet 4 | 78.98 |
| 3 | Claude Opus 4.8 | 78.38 |
| 4 | Claude Sonnet 4.6 | 77.9 |
| 5 | GPT-5.1 | 77.76 |
| 6 | Claude Sonnet 4.5 | 77.7 |
| 7 | GPT-5.5 | 77.33 |
| 8 | Claude Opus 4.7 | 76.97 |
| 9 | Grok 4.20 | 76.25 |
| 10 | GPT-5 | 76.24 |
| 11 | GPT-5.3 | 76.11 |
| 12 | GPT-5.2 | 75.87 |
| 13 | Mistral Medium 3.1 | 75.85 |
| 14 | GPT-4o | 75.82 |
| 15 | DeepSeek V3.2 Exp | 75.2 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-af · How It Works · Data refreshed daily, snapshot 2026-09-05.