Evals for Every Language - Language ki: leaderboard
Metric: Average Score (%). Source: huggingface.co. 70 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 43.55 |
| 2 | Gemini 2.5 Flash | 43.39 |
| 3 | Gemini 3.1 Pro (Preview) | 43.2 |
| 4 | Claude Sonnet 4.5 | 43.02 |
| 5 | Claude Opus 4.7 | 42.28 |
| 6 | Gemini 3.1 Flash Lite | 41.33 |
| 7 | Claude Sonnet 4 | 41.07 |
| 8 | GPT-5.3 | 39.05 |
| 9 | GPT-5.1 | 38.01 |
| 10 | GPT-5.5 | 37.87 |
| 11 | Claude Opus 4.8 | 36.93 |
| 12 | GPT-5.4 | 36.09 |
| 13 | Qwen 3.6 Plus | 35.95 |
| 14 | Grok 4.20 | 35.12 |
| 15 | GPT-5.2 | 34.47 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-ki · How It Works · Data refreshed daily, snapshot 2026-09-05.