Evals for Every Language - Language vai: leaderboard
Metric: Average Score (%). Source: huggingface.co. 61 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 33.33 |
| 2 | Mistral Small 3.2 | 25 |
| 3 | GPT-OSS-20B | 25 |
| 4 | DeepSeek V3.2 Exp | 25 |
| 5 | GPT-OSS-120B | 20 |
| 6 | Command A | 20 |
| 7 | Qwen 3.6 Flash | 12.5 |
| 8 | GPT-4.1 | 10 |
| 9 | GPT-5 Nano | 10 |
| 10 | Llama 4 Maverick | 10 |
| 11 | Mistral Medium 3.1 | 10 |
| 12 | GPT-5.3 | 10 |
| 13 | Gemma 3 27B (IT) | 0 |
| 14 | GPT-4o | 0 |
| 15 | Gemini 2.5 Pro | 0 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-vai · How It Works · Data refreshed daily, snapshot 2026-09-05.