Evals for Every Language - Language mfe: leaderboard
Metric: Average Score (%). Source: huggingface.co. 70 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 57.05 |
| 2 | Claude Sonnet 4.6 | 55.76 |
| 3 | GPT-5 | 55.55 |
| 4 | GPT-5.4 | 55.54 |
| 5 | GPT-5.1 | 54.82 |
| 6 | Gemini 2.5 Pro | 54.56 |
| 7 | Gemini 2.5 Flash | 54.54 |
| 8 | Claude Opus 4.7 | 54.32 |
| 9 | Claude Opus 4.8 | 53.48 |
| 10 | Claude Sonnet 4.5 | 52.35 |
| 11 | Grok 4.20 | 51.86 |
| 12 | Qwen 3.6 Plus | 51.57 |
| 13 | GPT-5.3 | 51.49 |
| 14 | GPT-5.2 | 50.87 |
| 15 | Gemini 3.1 Flash Lite | 50.68 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-mfe · How It Works · Data refreshed daily, snapshot 2026-09-05.