LogicKor - Grammar: leaderboard
LogicKor Korean-language grammar slice measuring grammatical knowledge and correction quality.
Metric: Score (0-10). Source: lk.instruct.kr. Saturation forecast: Estimated already saturated. 253 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3 Opus (20240229) | 9.64 | #412 |
| 2 | GPT-4o (2024-05-13) | 9.5 | #325 |
| 3 | Claude 3.5 Sonnet (20240620) | 9.21 | #299 |
| 4 | Gemini 1.5 Pro (001) | 9.21 | #444 |
| 5 | Gemini 1.5 Flash (001) | 8.85 | #577 |
| 6 | GPT-4 (0613) | 8.35 | #491 |
| 7 | Llama 3.1 405B Instruct FP8 | 8.35 | #389 |
| 8 | Claude 3 Sonnet (20240229) | 8.28 | #677 |
| 9 | GPT-4 Preview (0125) | 8.21 | #486 |
| 10 | GPT-4 Preview (1106) | 8.14 | #418 |
| 11 | Claude 3 Haiku (20240307) | 8.07 | #771 |
| 12 | GPT-3.5 Turbo (0125) | 8 | #984 |
| 13 | GPT-3.5 Turbo (1106) | 7.92 | #961 |
| 14 | Mistral Large 2 (Jul) | 7.78 | #484 |
| 15 | c4ai-command-r-plus | 7.78 | #724 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=logickor-grammar · How It Works · Data refreshed daily, snapshot 2026-10-20.