Dnotitia Korean LLM Leaderboard: leaderboard
Dnotitia's Korean LLM leaderboard: 30 private Korean items covering understanding, generation, reasoning, RAG and tool calling, each scored 0-1; the board is the published mean over repeated runs, shown as 0-100 (higher is better). The operator also builds DNA models, and the top is saturated (16 models at 98 or above).
Metric: Mean Item Score (%). Source: leaderboard.dnotitia.com. Saturation forecast: Estimated already saturated. 253 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 100 | #6 |
| 2 | Claude Opus 5.5 | 100 | #3 |
| 3 | Gemini 3.7 Flash | 100 | #24 |
| 4 | Gemini 4 Argon | 100 | #9 |
| 5 | Claude Fable 5.1 | 100 | #10 |
| 6 | Muse Spark 1.2 | 100 | #69 |
| 7 | Muse Spark 1.3 | 100 | #31 |
| 8 | Muse Spark Open | 100 | #42 |
| 9 | GPT-6.1 Sol | 99.82 | #7 |
| 10 | DeepSeek V4 Pro (0813) | 99.73 | #58 |
| 11 | Gemini 3.6 Flash | 99.55 | #57 |
| 12 | GPT-6 Sol | 99.55 | #19 |
| 13 | Grok 4.7 | 99.45 | #26 |
| 14 | Gemini 3.8 Flash | 99.36 | #30 |
| 15 | Qwen 3.8 Max | 99.29 | #34 |
Interactive version: theaggregate.ai/benchmark?slug=dnotitia-korean-llm-leaderboard · How It Works · Data refreshed daily, snapshot 2026-11-05.