Evals for Every Language - Translation To: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 41.34 |
| 2 | Gemini 2.5 Pro | 40.82 |
| 3 | Claude Opus 4.7 | 39.5 |
| 4 | GPT-5.5 | 39.31 |
| 5 | Gemini 2.5 Flash | 39.12 |
| 6 | Claude Opus 4.8 | 38.22 |
| 7 | Claude Sonnet 4.5 | 38.07 |
| 8 | Gemini 3.1 Flash Lite | 37.99 |
| 9 | GPT-5.4 | 37.74 |
| 10 | Qwen 3.6 Plus | 36.55 |
| 11 | Claude Sonnet 4 | 36.4 |
| 12 | GPT-5.3 | 36.27 |
| 13 | GPT-5.2 | 36.14 |
| 14 | Claude Sonnet 4.6 | 35.17 |
| 15 | Grok 4.20 | 34.82 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-translation-to · How It Works · Data refreshed daily, snapshot 2026-09-05.