Evals for Every Language - Language pt: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Flash Lite | 77.5 |
| 2 | Gemini 3.1 Pro (Preview) | 75.55 |
| 3 | Claude Sonnet 4.5 | 75.39 |
| 4 | Claude Opus 4.8 | 75.31 |
| 5 | Claude Sonnet 4.6 | 74.93 |
| 6 | Claude Opus 4.7 | 74.77 |
| 7 | Claude Sonnet 4 | 74.42 |
| 8 | Gemini 2.5 Pro | 74.29 |
| 9 | GPT-5.3 | 74.22 |
| 10 | Gemini 2.5 Flash Lite | 73.59 |
| 11 | Gemma 3 27B (IT) | 73.25 |
| 12 | GPT-5.5 | 73.02 |
| 13 | GPT-5 | 72.95 |
| 14 | Command A | 72.95 |
| 15 | Claude Haiku 4.5 | 72.61 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-pt · How It Works · Data refreshed daily, snapshot 2026-09-05.