Evals for Every Language - Language am: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Flash Lite | 68.6 |
| 2 | Claude Opus 4.7 | 67.86 |
| 3 | Gemini 3.1 Pro (Preview) | 67.7 |
| 4 | Claude Opus 4.8 | 65.76 |
| 5 | GPT-5.5 | 65.01 |
| 6 | Grok 4.20 | 64.62 |
| 7 | Claude Sonnet 4.6 | 64.23 |
| 8 | GPT-5.2 | 63.32 |
| 9 | Qwen 3.6 Plus | 62.64 |
| 10 | DeepSeek V3.1 | 62.28 |
| 11 | GPT-5.3 | 61.75 |
| 12 | GPT-5.4 | 61.3 |
| 13 | Claude Sonnet 4.5 | 59.53 |
| 14 | Claude Sonnet 4 | 57.5 |
| 15 | Qwen 3.6 Flash | 57.31 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-am · How It Works · Data refreshed daily, snapshot 2026-09-05.