Evals for Every Language - Language tg: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 69.8 |
| 2 | GPT-5.1 | 69.31 |
| 3 | Claude Opus 4.7 | 68.9 |
| 4 | Gemini 3.1 Flash Lite | 68.81 |
| 5 | GPT-5 | 68.67 |
| 6 | Claude Opus 4.8 | 68.57 |
| 7 | Claude Sonnet 4.6 | 68.5 |
| 8 | GPT-5.4 | 68.11 |
| 9 | GPT-5.5 | 67.17 |
| 10 | Claude Sonnet 4.5 | 67.09 |
| 11 | GPT-5.3 | 66.59 |
| 12 | GPT-4o | 66.32 |
| 13 | Qwen 3.6 Plus | 66.31 |
| 14 | Claude Sonnet 4 | 66.15 |
| 15 | Grok 4.20 | 65.11 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-tg · How It Works · Data refreshed daily, snapshot 2026-09-05.