Benchmark Almanac

LegalCiteBench

Claude Sonnet 4.5 0.78 leads since February 2026
Not in model cards

Models recover and verify citations from United States judicial opinions without external lookup. The benchmark also tests whether models match cases and correct citation errors.

Made by
Sijia Chen and colleagues
Released
February 2026
Kind
Law
Level
University
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Subject
Law
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Law and finance
Strongest model tested
Claude Sonnet 4.5 #81 on the capability index
Last model added
October 2025
Scores refreshed
November 20th, 2026
21
Entries on its own leaderboard, 14 models we track

Claude Sonnet 4.5 leads with 0.78 among models released by October 2025 Claude Sonnet 4.5 leads with 0.78 among models released by October 2025

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in October 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Claude Sonnet 4.5 0.78
  2. 2 Equall/SaulLM-54B-Instruct 0.72
  3. 3 Qwen3-14B 0.72
  4. 4 DeepSeek-V3.1 0.70
  5. 5 o4-mini 0.70
  6. 6 Gemini 2.5 Flash Sep 2025 0.69
  7. 7 meta-llama/Llama-3.1-70B-Instruct 0.64
  8. 8 GPT-5 mini 0.64
  9. 9 DeepSeek-R1-Distill-Qwen-1.5B 0.63
  10. 10 Qwen3-30B-A3B-Instruct Jul 2025 0.63
  11. 11 Claude Haiku 4.5 0.62
  12. 12 DeepSeek-R1-Distill-Qwen-7B 0.61
  13. 13 qwen3-4b-instruct-2507 0.59
  14. 14 mistralai/Mistral-7B-Instruct-v0.3 0.58
  15. 15 Phi-4 0.58
  16. 16 Qwen/Qwen2.5-7B-Instruct 0.55
  17. 17 Gemma 3 12B 0.51
  18. 18 GPT-4o mini 0.50
  19. 19 meta-llama/Llama-3.1-8B-Instruct 0.50
  20. 20 meta-llama/Llama-3.2-3B-Instruct 0.43
  21. 21 meta-llama/Llama-3.2-1B-Instruct 0.39

Source: Official LegalCiteBench leaderboard

#EntryScoreSettingDateSource
1 Claude Sonnet 4.5 0.78 Sep 2025 Official leaderboard
2 Equall/SaulLM-54B-Instruct 0.72 Official leaderboard
3 Qwen3-14B 0.72 Apr 2025 Official leaderboard
4 DeepSeek-V3.1 0.70 Aug 2025 Official leaderboard
5 o4-mini 0.70 Apr 2025 Official leaderboard
6 Gemini 2.5 Flash Sep 2025 0.69 Sep 2025 Official leaderboard
7 meta-llama/Llama-3.1-70B-Instruct open weights 0.64 Official leaderboard
8 GPT-5 mini 0.64 Aug 2025 Official leaderboard
9 DeepSeek-R1-Distill-Qwen-1.5B 0.63 Jan 2025 Official leaderboard
10 Qwen3-30B-A3B-Instruct Jul 2025 0.63 Jul 2025 Official leaderboard
11 Claude Haiku 4.5 0.62 Oct 2025 Official leaderboard
12 DeepSeek-R1-Distill-Qwen-7B 0.61 Jan 2025 Official leaderboard
13 qwen3-4b-instruct-2507 0.59 Aug 2025 Official leaderboard
14 mistralai/Mistral-7B-Instruct-v0.3 open weights 0.58 Official leaderboard
15 Phi-4 0.58 Dec 2024 Official leaderboard
16 Qwen/Qwen2.5-7B-Instruct open weights 0.55 Official leaderboard
17 Gemma 3 12B 0.51 Mar 2025 Official leaderboard
18 GPT-4o mini 0.50 Jul 2024 Official leaderboard
19 meta-llama/Llama-3.1-8B-Instruct open weights 0.50 Official leaderboard
20 meta-llama/Llama-3.2-3B-Instruct open weights 0.43 Official leaderboard
21 meta-llama/Llama-3.2-1B-Instruct open weights 0.39 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Sonnet 4.5 0.78 Sep 2025 Official leaderboard
2 Qwen3-14B 0.72 Apr 2025 Official leaderboard
3 DeepSeek-V3.1 0.70 Aug 2025 Official leaderboard
4 o4-mini 0.70 Apr 2025 Official leaderboard
5 Gemini 2.5 Flash Sep 2025 0.69 Sep 2025 Official leaderboard
6 GPT-5 mini 0.64 Aug 2025 Official leaderboard
7 DeepSeek-R1-Distill-Qwen-1.5B 0.63 Jan 2025 Official leaderboard
8 Qwen3-30B-A3B-Instruct Jul 2025 0.63 Jul 2025 Official leaderboard
9 Claude Haiku 4.5 0.62 Oct 2025 Official leaderboard
10 DeepSeek-R1-Distill-Qwen-7B 0.61 Jan 2025 Official leaderboard
11 qwen3-4b-instruct-2507 0.59 Aug 2025 Official leaderboard
12 Phi-4 0.58 Dec 2024 Official leaderboard
13 Gemma 3 12B 0.51 Mar 2025 Official leaderboard
14 GPT-4o mini 0.50 Jul 2024 Official leaderboard

Rankings by part

We read 5 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Citation Retrieval, Citation Completion and Citation Error Detection. Claude Sonnet 4.5 leads 2 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

Citation Retrieval

14 models we track have a score here.

  1. 1 Claude Sonnet 4.5 6.80
  2. 2 o4-mini ↑2 5.73
  3. 3 DeepSeek-V3.1 5.31
  4. 4 Gemini 2.5 Flash Sep 2025 ↑1 5.20
  5. 5 Claude Haiku 4.5 ↑4 4.42
  6. 6 Qwen3-30B-A3B-Instruct Jul 2025 ↑2 3.65
  7. 7 Phi-4 ↑5 3.59
  8. 8 qwen3-4b-instruct-2507 ↑3 3.33
  9. 9 Qwen3-14B ↓7 2.54
  10. 10 GPT-4o mini ↑4 2.40

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Own leaderboard Not tested since Feb 2026

Source: Official LegalCiteBench leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard
0.50 0.60 0.70 0.80 2025 0.50 0.60 0.70 0.80 2025

Source: Official LegalCiteBench leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Law, finance and medicine

All 637 in Law, finance and medicine
  • MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy. Qwen 3.8 Max 80.4%
  • HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
  • HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
  • FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy. Ling-3.1-flash 81.7%
  • KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. o1-preview 92.7%
  • Harvey LAB-AA v1.1 In use Artificial Analysis's run of Harvey's Legal Agent Benchmark on Harvey's newer private set of 120 tasks across 24 practice areas. A deliverable with a material hallucination scores zero, and the headline is the share of judges who find every rubric criterion passed. GPT-6.2 Astra 15%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 13 months ago, in October 2025, for Claude Haiku 4.5. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (21 entries).
  • Pass Models scored 14 models we track, plus 7 other entries on its leaderboard.
  • Warning Newest score The newest came 13 months ago, in October 2025, for Claude Haiku 4.5.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude Sonnet 4.5 (#81).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on huggingface.co: 14 rows for models we track and 7 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check