Evals for Every Language - Korean (ko): leaderboard
Average score across evaluation tasks in Korean (ko), from Evals for Every Language.
Metric: Average Score (%). Source: huggingface.co. Saturation forecast: Around February 2027. 68 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 77.2 | #58 |
| 2 | Claude Sonnet 4 | 76.48 | #199 |
| 3 | Claude Sonnet 4.5 | 75.63 | #142 |
| 4 | GPT-5.1 | 74.63 | #135 |
| 5 | Claude Opus 4.8 | 74.44 | #37 |
| 6 | Claude Sonnet 4.6 | 74.26 | #90 |
| 7 | GPT-5 | 72.78 | #98 |
| 8 | GPT-5.3 | 72.68 | #103 |
| 9 | Grok 4.20 | 72.64 | #175 |
| 10 | GPT-5.5 | 72.3 | #31 |
| 11 | Claude Opus 4.7 | 71.8 | #49 |
| 12 | GPT-5.2 | 71.59 | #110 |
| 13 | GPT-4o | 71.23 | #343 |
| 14 | Gemini 2.5 Pro | 70.74 | #152 |
| 15 | GPT-5.4 | 70.01 | #79 |
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-ko · How It Works · Data refreshed daily, snapshot 2026-10-20.