Evals for Every Language - Language dar — leaderboard
Metric: Average Score (%). Source: huggingface.co. 70 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 43.4 |
| 2 | Gemini 3.1 Pro (Preview) | 42.54 |
| 3 | Gemini 2.5 Pro | 42.49 |
| 4 | Gemini 3.1 Flash Lite | 40.6 |
| 5 | Claude Sonnet 4.5 | 38.66 |
| 6 | Gemini 2.5 Flash | 36.84 |
| 7 | GPT-5.4 | 36.04 |
| 8 | Claude Sonnet 4.6 | 34.18 |
| 9 | GPT-5.5 | 34.18 |
| 10 | Claude Opus 4.8 | 33.01 |
| 11 | GPT-5 | 32.21 |
| 12 | Claude Sonnet 4 | 32 |
| 13 | GPT-5.2 | 31.2 |
| 14 | Mistral Medium 3.1 | 31.04 |
| 15 | Qwen 3.6 Plus | 30.35 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-dar · How the rankings work · Data refreshed daily, snapshot 2026-07-22.