Legal Research Bench: leaderboard
Metric: Score (%). Source: benchmarklist.com. 27 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 5 | 55.29 |
| 2 | Claude Fable 5 | 49.52 |
| 3 | GPT-5.6 Sol | 48.08 |
| 4 | Kimi K3 | 44.23 |
| 5 | Claude Opus 4.8 | 43.75 |
| 6 | Claude Sonnet 5 | 41.83 |
| 7 | GPT-5.6 Terra | 40.87 |
| 8 | GPT-5.5 | 40.39 |
| 9 | Claude Sonnet 4.6 | 38.46 |
| 10 | Grok 4.5 | 37.98 |
| 11 | Muse Spark 1.1 | 37.98 |
| 12 | GPT-5.6 Luna | 36.54 |
| 13 | GLM-5.2 | 31.25 |
| 14 | Gemini 3.5 Flash | 30.77 |
| 15 | MiniMax-M3 | 29.81 |
Interactive version: theaggregate.ai/benchmark?slug=legal-research-bench · How It Works · Data refreshed daily, snapshot 2026-09-19.