Evals for Every Language - Language tpi — leaderboard
Metric: Average Score (%). Source: huggingface.co. 70 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 51.03 |
| 2 | Claude Opus 4.7 | 50.59 |
| 3 | Gemini 2.5 Pro | 50.45 |
| 4 | Claude Opus 4.8 | 50.19 |
| 5 | Gemini 3.1 Flash Lite | 48.57 |
| 6 | Claude Sonnet 4 | 48.48 |
| 7 | Grok 4.20 | 47.43 |
| 8 | GPT-4o | 47.18 |
| 9 | Claude Sonnet 4.5 | 47.17 |
| 10 | GPT-5.1 | 47.15 |
| 11 | Gemini 2.5 Flash | 47.11 |
| 12 | GPT-5 Mini | 46.38 |
| 13 | Kimi K2 | 46.05 |
| 14 | GPT-5 | 45.94 |
| 15 | GPT-5.5 | 45.75 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-tpi · How the rankings work · Data refreshed daily, snapshot 2026-07-22.