Evals for Every Language - Language nso: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 71.26 |
| 2 | GPT-5.5 | 68.22 |
| 3 | Claude Opus 4.7 | 66.81 |
| 4 | GPT-5.3 | 66.32 |
| 5 | Claude Opus 4.8 | 65.26 |
| 6 | Gemini 3.1 Flash Lite | 64.74 |
| 7 | GPT-5.4 | 64.64 |
| 8 | DeepSeek V3.1 | 61.81 |
| 9 | GPT-5.2 | 60.77 |
| 10 | Grok 4.20 | 60.31 |
| 11 | Claude Sonnet 4.6 | 58.9 |
| 12 | Qwen 3.6 Plus | 58.43 |
| 13 | Gemini 2.5 Pro | 58.08 |
| 14 | Claude Sonnet 4.5 | 57.16 |
| 15 | GPT-5 | 56.47 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-nso · How It Works · Data refreshed daily, snapshot 2026-09-05.