Korean Canonical Legal Benchmark: leaderboard
Korean legal benchmark with multiple-choice and essay-style reasoning tasks, including support for precedent-based answers.
Metric: Mean with supporting precedents (imported). Source: arxiv.org. Source: benchmarklist.com. Saturation forecast: Around November 2026. 31 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 82.1 | #152 |
| 2 | GPT-5 | 78.5 | #98 |
| 3 | Gemini 2.5 Flash | 77.4 | #245 |
| 4 | Claude Sonnet 4.5 | 76.85 | #142 |
| 5 | O3 | 76.4 | #124 |
| 6 | Claude 3.7 Sonnet | 72.8 | #249 |
| 7 | Claude Opus 4.1 | 71.8 | #140 |
| 8 | GPT-5 Mini | 71.15 | #186 |
| 9 | Qwen 3 235B A22B (Thinking) | 70.4 | #313 (Qwen 3 235B A22B) |
| 10 | GPT-4.1 | 68.55 | #248 |
| 11 | Claude Sonnet 4 | 66.75 | #199 |
| 12 | O4 Mini | 64.55 | #180 |
| 13 | Kimi K2 | 63.65 | #243 |
| 14 | DeepSeek V3 | 63.45 | #321 |
| 15 | DeepSeek R1 | 62.75 | #251 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=korean-canonical-legal-benchmark · How It Works · Data refreshed daily, snapshot 2026-10-20.