LogicKor - Multi-Turn: leaderboard
LogicKor multi-turn aggregate measuring Korean prompt performance across conversational context.
Metric: Score (0-10). Source: lk.instruct.kr. Saturation forecast: Estimated already saturated. 253 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o (2024-05-13) | 9.5 | #325 |
| 2 | Claude 3.5 Sonnet (20240620) | 9.47 | #299 |
| 3 | Gemini 1.5 Pro (001) | 9.19 | #444 |
| 4 | Mistral Large 2 (Jul) | 9.16 | #484 |
| 5 | GPT-4 Preview (1106) | 9.07 | #418 |
| 6 | Llama 3.1 405B Instruct FP8 | 9.04 | #389 |
| 7 | Claude 3 Opus (20240229) | 8.97 | #412 |
| 8 | Gemini 1.5 Flash (001) | 8.95 | #577 |
| 9 | GPT-4 Preview (0125) | 8.92 | #486 |
| 10 | ko-gemma-2-9B-it | 8.76 | #541 |
| 11 | GPT-4 Turbo | 8.66 | #404 |
| 12 | GPT-4 (0613) | 8.61 | #491 |
| 13 | Qwen 2 72B Instruct | 8.47 | #576 |
| 14 | Claude 3 Sonnet (20240229) | 8.4 | #677 |
| 15 | Claude 3 Haiku (20240307) | 8.35 | #771 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=logickor-multi-turn · How It Works · Data refreshed daily, snapshot 2026-10-20.