DLawBench - US Law: leaderboard
Metric: Resolution score (0-100), a rate scaled by 100: mean of fact resolution (annotated facts carried into the legal memo and correctly preserved or reframed against the court record) and issue resolution (expert-specified legal analysis points addressed), U.S.-law cases only, on 461 real Chinese-law and U.S.-law cases replayed as multi-turn consultations with a Claude Sonnet 4.6 client simulator in four narrative styles; judged by a GPT-5.1, Claude Opus 4.6 and Gemini 3.1 Pro panel (median; same-vendor judges recuse), empty memos count as zero; provider decoding defaults; higher is better. Source: arxiv.org. Saturation forecast: Around September 2027. 26 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 51.2 |
| 2 | GPT-5.4 | 49.7 |
| 3 | GPT-5.2 | 44.4 |
| 4 | Gemini 3.1 Pro (Preview) | 43.3 |
| 5 | Claude Opus 4.6 | 42.9 |
| 6 | Claude Opus 4.7 | 41.2 |
| 7 | DeepSeek V4 Pro | 39.4 |
| 8 | GLM-5.1 | 39.3 |
| 9 | Qwen 3.6 Max Preview | 39.1 |
| 10 | Claude Sonnet 4.6 | 37.8 |
| 11 | Grok 4.1 Fast | 35.9 |
| 12 | Kimi K2.6 | 34.7 |
| 13 | GLM-5 | 34.2 |
| 14 | Kimi K2.5 | 33.4 |
| 15 | Qwen 3.6 Plus | 32.2 |
Interactive version: theaggregate.ai/benchmark?slug=dlawbench-us-law · How It Works · Data refreshed daily, snapshot 2026-09-29.