Evals for Every Language - Language yi: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Flash Lite | 72.51 |
| 2 | Gemini 3.1 Pro (Preview) | 71 |
| 3 | Claude Sonnet 4.5 | 70.62 |
| 4 | Claude Opus 4.7 | 69.32 |
| 5 | Claude Sonnet 4 | 68.85 |
| 6 | Claude Sonnet 4.6 | 67.99 |
| 7 | GPT-4o | 67.87 |
| 8 | GPT-5 | 67.82 |
| 9 | Claude Opus 4.8 | 67.09 |
| 10 | GLM 4.5 Air | 66.9 |
| 11 | GPT-5.3 | 66.81 |
| 12 | GPT-5.1 | 66.59 |
| 13 | GPT-5.2 | 65.93 |
| 14 | Nova Pro (v1) | 65.1 |
| 15 | GPT-5.5 | 65.01 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-yi · How It Works · Data refreshed daily, snapshot 2026-09-05.