Evals for Every Language - Language he: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.5 | 77 |
| 2 | Gemini 3.1 Pro (Preview) | 76.35 |
| 3 | Claude Opus 4.7 | 74.47 |
| 4 | Gemini 3.1 Flash Lite | 73.92 |
| 5 | Claude Sonnet 4.6 | 73.79 |
| 6 | Gemini 2.5 Pro | 73.5 |
| 7 | Claude Sonnet 4 | 73.15 |
| 8 | GPT-5.1 | 72.55 |
| 9 | DeepSeek V3.1 | 72.44 |
| 10 | Claude Opus 4.8 | 71.83 |
| 11 | DeepSeek V3.2 Exp | 71.58 |
| 12 | Gemma 3 27B (IT) | 71.57 |
| 13 | GPT-5.3 | 71.18 |
| 14 | Grok 4.20 | 71 |
| 15 | Qwen 3.6 Plus | 70.22 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-he · How It Works · Data refreshed daily, snapshot 2026-09-05.