Evals for Every Language - Language kea: leaderboard
Metric: Average Score (%). Source: huggingface.co. 70 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 62.21 |
| 2 | Gemini 3.1 Pro (Preview) | 61.08 |
| 3 | Gemini 2.5 Flash | 58.07 |
| 4 | Claude Sonnet 4.5 | 56.3 |
| 5 | GPT-5.4 | 55.33 |
| 6 | Claude Opus 4.7 | 54.47 |
| 7 | Gemini 3.1 Flash Lite | 53.5 |
| 8 | GPT-5 | 53.13 |
| 9 | GPT-4o | 52.85 |
| 10 | Claude Sonnet 4 | 52.85 |
| 11 | GPT-5.5 | 52.65 |
| 12 | GPT-5.1 | 51.98 |
| 13 | Claude Opus 4.8 | 51.91 |
| 14 | GPT-5.3 | 51.39 |
| 15 | Grok 4.20 | 50.26 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-kea · How It Works · Data refreshed daily, snapshot 2026-09-05.