legal benchmark
Legal Agent Benchmark Leaderboard
Legal Agent Benchmark (LAB) is Harvey's open-source benchmark for evaluating AI agents on complex, long-horizon legal work.
Updated Sep 5, 2026
Models13
Model coverage13
MetricScore
EvidenceB
Legal Agent Benchmark Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAN | Score13.30% | Percentile100.00% | Participants13 | EvidenceC | Evaluated |
| Rank02 | ModelAN | Score11.70% | Percentile91.67% | Participants13 | EvidenceC | Evaluated |
| Rank03 | ModelAN | Score7.10% | Percentile83.33% | Participants13 | EvidenceB | Evaluated |
| Rank04 | ModelAN | Score5.80% | Percentile75.00% | Participants13 | EvidenceC | Evaluated |
| Rank05 | ModelAN | Score5.40% | Percentile66.67% | Participants13 | EvidenceB | Evaluated |
| Rank06 | ModelAN | Score4.20% | Percentile58.33% | Participants13 | EvidenceB | Evaluated |
| Rank07 | ModelOP | Score2.10% | Percentile50.00% | Participants13 | EvidenceB | Evaluated |
| Rank08 | ModelGO | Score0.80% | Percentile41.67% | Participants13 | EvidenceB | Evaluated |
| Rank09 | ModelOP | Score0.40% | Percentile33.33% | Participants13 | EvidenceB | Evaluated |
| Rank10 | ModelGO | Score0.00% | Percentile25.00% | Participants13 | EvidenceB | Evaluated |
| Rank11 | ModelGO | Score0.00% | Percentile16.67% | Participants13 | EvidenceB | Evaluated |
| Rank12 | ModelGO | Score0.00% | Percentile8.33% | Participants13 | EvidenceB | Evaluated |
| Rank13 | ModelOP | Score0.00% | Percentile0.00% | Participants13 | EvidenceB | Evaluated |
Legal Agent Benchmark Highlights
The leading models and scores on this benchmark.
Legal Agent Benchmark Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for Legal Agent Benchmark
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is Legal Agent Benchmark?
What Legal Agent Benchmark measures and how its scores work.
Legal Agent Benchmark (LAB) is a benchmark for evaluating AI agents on complex, long-horizon legal work.
LAB reports a Score as a ratio using an all-pass standard against expert-curated rubrics; a task passes only if every required criterion for facts, conclusions, citations, structure, and analytical moves passes.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- Legal Agent Benchmark
- Modality
- text
- Primary category
- legal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about Legal Agent Benchmark.
Which model scores highest on Legal Agent Benchmark?
Claude Fable 5 is currently ranked first with 13.30%.
What are the top three models on Legal Agent Benchmark?
The current leaders are Claude Fable 5 (13.30%), Claude Opus 5 (11.70%), and Claude Opus 4.7 (7.10%).
Which Legal Agent Benchmark model has the lowest official input price?
Gemini 3.1 Flash-Lite has the lowest matched official input price at $0.25 input / $1.5 output per 1M tokens.
Which models are fastest among Legal Agent Benchmark results?
The fastest matched records are Gemini 3.5 Flash (210.94 tok/s via Google), GPT-5.5 (134.94 tok/s via OpenAI), and Gemini 3 Flash (124.32 tok/s via Google).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.