Evals for Every Language - Language vec: leaderboard
Metric: Average Score (%). Source: huggingface.co. 70 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 54.06 |
| 2 | Gemini 3.1 Flash Lite | 53.71 |
| 3 | Gemini 3.1 Pro (Preview) | 53.09 |
| 4 | Gemini 2.5 Flash | 53.09 |
| 5 | Claude Opus 4.7 | 52.38 |
| 6 | GPT-5 | 51.65 |
| 7 | Claude Sonnet 4.6 | 51.48 |
| 8 | Claude Sonnet 4.5 | 51.1 |
| 9 | Claude Sonnet 4 | 51.01 |
| 10 | GPT-5.5 | 51 |
| 11 | GPT-5.3 | 50.57 |
| 12 | GPT-5.1 | 50.4 |
| 13 | Grok 4.20 | 50.27 |
| 14 | GPT-4o | 49.49 |
| 15 | GPT-5.4 | 49.09 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-vec · How It Works · Data refreshed daily, snapshot 2026-09-05.