Evals for Every Language - Language awa: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 69.39 |
| 2 | Claude Opus 4.7 | 68.23 |
| 3 | Gemini 2.5 Pro | 68.19 |
| 4 | Claude Opus 4.8 | 67.71 |
| 5 | Claude Sonnet 4.5 | 67.7 |
| 6 | Gemini 3.1 Flash Lite | 67.13 |
| 7 | GPT-5.3 | 66.44 |
| 8 | GPT-5.5 | 66.14 |
| 9 | GPT-5 | 66.02 |
| 10 | GPT-5.2 | 65.9 |
| 11 | Claude Sonnet 4 | 65.53 |
| 12 | Nova Pro (v1) | 65.46 |
| 13 | Mistral Medium 3.1 | 65.19 |
| 14 | GPT-5.1 | 64.74 |
| 15 | Gemma 3 27B (IT) | 64.64 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-awa · How It Works · Data refreshed daily, snapshot 2026-09-05.