JudgmentBench
Models produce legal work for 30 tasks. Practicing lawyers grade individual answers and compare pairs of answers.
- Made by
- Stanford Law School, Snorkel AI, and Harvey
- Released
- May 2026
- Kind
- Law
- Level
- Expert
- Scored
- Graded by people
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Subject
- Law
- Language
- English
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- Occupations
- Legal
- In views
- Law and finance
- Strongest model tested
- Claude Opus 4.8 #22 on the capability index
- Last model added
- July 2026
- Scores refreshed
- November 20th, 2026
Claude Opus 4.8 leads with 89% among models released by July 2026
Best score per model from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.
The newest model with a score here came out in July 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 Claude Opus 4.8 Comparative judgment 89%
- 2 GLM-5.2 Comparative judgment 87.5%
- 3 Claude Sonnet 5 Comparative judgment 85.2%
- 4
Qwen3.7-Max Comparative judgment 84.1%
- 5
Gemini 3.1 Pro Comparative judgment 82%
- 6
Gemini 3.6 Flash Comparative judgment 81.9%
- 7 GPT-5.4 Mini Comparative judgment 70.2%
- 8 GPT-5.4 Comparative judgment 59.9%
Source: The JudgmentBench paper
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.8 | 89% | Comparative judgment | May 2026 | Official leaderboard |
| 2 | GLM-5.2 | 87.5% | Comparative judgment | Jun 2026 | Official leaderboard |
| 3 | Claude Sonnet 5 | 85.2% | Comparative judgment | Jun 2026 | Official leaderboard |
| 4 | | 84.1% | Comparative judgment | May 2026 | Official leaderboard |
| 5 | | 82% | Comparative judgment | Feb 2026 | Official leaderboard |
| 6 | | 81.9% | Comparative judgment | Jul 2026 | Official leaderboard |
| 7 | GPT-5.4 Mini | 70.2% | Comparative judgment | Mar 2026 | Official leaderboard |
| 8 | GPT-5.4 | 59.9% | Comparative judgment | Mar 2026 | Official leaderboard |
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: The JudgmentBench paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Claude Opus 4.8 | Comparative judgment | 89% | |||
| GLM-5.2 | Comparative judgment | 87.5% | |||
| Claude Sonnet 5 | Comparative judgment | 85.2% | |||
| Comparative judgment | 84.1% | ||||
| Comparative judgment | 82% | ||||
| Comparative judgment | 81.9% | ||||
| GPT-5.4 Mini | Comparative judgment | 70.2% | |||
| GPT-5.4 | Comparative judgment | 59.9% | |||
| Rubric scoring | 56% | ||||
| Rubric scoring | 53.3% | ||||
| Rubric scoring | 52.4% | ||||
| GLM-5.2 | Rubric scoring | 48.4% | |||
| Claude Opus 4.8 | Rubric scoring | 46.5% | |||
| GPT-5.4 Mini | Rubric scoring | 44.9% | |||
| Claude Sonnet 5 | Rubric scoring | 44.6% | |||
| GPT-5.4 | Rubric scoring | 42.7% |
3 models have held number one in the benchmark’s paper since its release in May 2026
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The JudgmentBench paper
Claude Opus 4.8 on the official leaderboard passed 85% in May 2026
Models released before it came out already scored that high. All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official JudgmentBench 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
- 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
All 9 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the benchmark’s paper (8 entries).
- Pass Models scored 8 models we track.
- Pass Newest score The newest is from July 2026, for Gemini 3.6 Flash.
- Note Frontier models None of the top 10 has a score, which is fine for a retired test. The strongest tested is Claude Opus 4.8 (#22).
- Note Headroom Saturated. The best is 89%, past 85% since May 2026.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 16 rows for models we track 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.