Evals for Every Language - Language sm: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.8 | 72.36 |
| 2 | Gemini 3.1 Pro (Preview) | 71.3 |
| 3 | GPT-5.5 | 70.86 |
| 4 | GPT-5 | 68.17 |
| 5 | Claude Opus 4.7 | 68.11 |
| 6 | GPT-5.3 | 67.84 |
| 7 | Gemini 2.5 Pro | 67.8 |
| 8 | Claude Sonnet 4.6 | 67.3 |
| 9 | GPT-5.4 | 67.25 |
| 10 | GPT-5.1 | 67.06 |
| 11 | Grok 4.20 | 66.58 |
| 12 | Gemini 3.1 Flash Lite | 64.92 |
| 13 | GPT-4o | 64.37 |
| 14 | Qwen 3.6 Plus | 64.11 |
| 15 | Claude Sonnet 4.5 | 63.1 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-sm · How It Works · Data refreshed daily, snapshot 2026-09-05.