Evals for Every Language - Language ti: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 66.67 |
| 2 | Claude Opus 4.7 | 64.96 |
| 3 | Gemini 3.1 Flash Lite | 63.4 |
| 4 | GPT-5.5 | 62.92 |
| 5 | Claude Sonnet 4.6 | 62.59 |
| 6 | Claude Opus 4.8 | 62.02 |
| 7 | GPT-5.4 | 61.53 |
| 8 | Qwen 3.6 Plus | 61.19 |
| 9 | GPT-5.3 | 59.93 |
| 10 | Claude Sonnet 4.5 | 59.64 |
| 11 | Grok 4.20 | 58.99 |
| 12 | GPT-5.2 | 58.86 |
| 13 | Claude Sonnet 4 | 57.36 |
| 14 | GPT-5 | 56.78 |
| 15 | Gemini 2.5 Pro | 56.42 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-ti · How It Works · Data refreshed daily, snapshot 2026-09-05.