Evals for Every Language - Language sd: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 71.63 |
| 2 | Gemini 3.1 Pro (Preview) | 71.51 |
| 3 | Gemini 2.5 Pro | 71.03 |
| 4 | Gemini 3.1 Flash Lite | 70.74 |
| 5 | Claude Sonnet 4.5 | 70.49 |
| 6 | GPT-5.1 | 70.12 |
| 7 | GPT-5.3 | 69.99 |
| 8 | GPT-4o | 69.14 |
| 9 | Claude Sonnet 4 | 69 |
| 10 | Qwen 3.6 Plus | 68.91 |
| 11 | GPT-5.4 | 68.88 |
| 12 | GPT-5.5 | 68.86 |
| 13 | Claude Opus 4.8 | 68.48 |
| 14 | Claude Sonnet 4.6 | 68.42 |
| 15 | GPT-5 | 68.13 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-sd · How It Works · Data refreshed daily, snapshot 2026-09-05.