LegalCiteBench - Citation Retrieval: leaderboard
Metric: Cat1 citation-level F1 (0-100) between the cited authorities the model lists for an opinion-derived legal research question and the reporter citations of the source opinion (4,899 instances), closed-book (no retrieval) on instances derived from 1,000 U.S. judicial opinions of the Case Law Access Project, greedy decoding at temperature 0 with at most 400 generated tokens, GPT-4o-mini judge; higher is better. Source: arxiv.org. Saturation forecast: Around 2031. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.5 | 6.8 |
| 2 | O4 Mini | 5.73 |
| 3 | DeepSeek V3.1 | 5.31 |
| 4 | Gemini 2.5 Flash | 5.2 |
| 5 | Claude Haiku 4.5 | 4.42 |
| 6 | Llama 3.1 70B | 3.82 |
| 7 | Qwen 3 30B A3B | 3.65 |
| 8 | Phi-4 | 3.59 |
| 9 | Qwen 3 4B | 3.33 |
| 10 | Mistral 7B | 2.85 |
| 11 | Qwen 3 14B | 2.54 |
| 12 | GPT-4o Mini | 2.4 |
| 13 | Gemma 3 12B | 1.92 |
| 14 | GPT-5 Mini | 1.81 |
| 15 | Llama 3.1 8B | 1.47 |
Interactive version: theaggregate.ai/benchmark?slug=legalcitebench-citation-retrieval · How It Works · Data refreshed daily, snapshot 2026-10-07.