Evals for Every Language - Shan (shn): leaderboard
Average score across evaluation tasks in Shan (shn), from Evals for Every Language.
Metric: Average Score (%). Source: huggingface.co. Saturation forecast: Around July 2027. 61 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.8 | 68.4 | #37 |
| 2 | Claude Opus 4.7 | 65.62 | #49 |
| 3 | GPT-5.5 | 64.19 | #31 |
| 4 | Gemini 3.1 Pro (Preview) | 63.72 | #58 |
| 5 | Qwen 3.6 Plus | 56.61 | #120 |
| 6 | Gemini 3.1 Flash Lite | 55.47 | #191 |
| 7 | Claude Sonnet 4.6 | 53.4 | #90 |
| 8 | GPT-5.4 | 52.19 | #79 |
| 9 | Gemini 2.5 Pro | 50.98 | #152 |
| 10 | GPT-5.3 | 50.6 | #103 |
| 11 | DeepSeek V3.2 Exp | 49.41 | #234 |
| 12 | GPT-5.2 | 49.15 | #110 |
| 13 | Qwen 3.6 Flash | 46.16 | #214 |
| 14 | Nova Pro (v1) | 41.11 | #497 |
| 15 | Gemma 3 27B (IT) | 40.15 | #523 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-shn · How It Works · Data refreshed daily, snapshot 2026-10-20.