Evals for Every Language - Language ru: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 75.78 |
| 2 | Gemini 2.5 Pro | 74.8 |
| 3 | Gemini 3.1 Flash Lite | 73.42 |
| 4 | Claude Sonnet 4.6 | 73.17 |
| 5 | Claude Opus 4.8 | 72.92 |
| 6 | Claude Sonnet 4 | 72.49 |
| 7 | GPT-5.3 | 72.47 |
| 8 | Claude Sonnet 4.5 | 72.31 |
| 9 | GPT-5.1 | 71.92 |
| 10 | DeepSeek V3.2 Exp | 71.89 |
| 11 | Llama 4 Maverick | 71.63 |
| 12 | GPT-5.2 | 71.56 |
| 13 | GPT-5.4 | 71.2 |
| 14 | Claude Opus 4.7 | 71.09 |
| 15 | Grok 4.20 | 70.96 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-ru · How It Works · Data refreshed daily, snapshot 2026-09-05.