cd /news/artificial-intelligence/harvey-lab-aa-evaluating-ai-agents-o… · home topics artificial-intelligence article
[ARTICLE · art-58411] src=artificialanalysis.ai ↗ pub= topic=artificial-intelligence verified=true sentiment=· neutral

Harvey LAB-AA: evaluating AI agents on real-world legal work

Harvey LAB-AA, a new legal agent benchmark from Artificial Analysis, evaluates AI models on 120 real-world legal tasks across 24 practice areas. Claude Fable 5 leads with a 14.2% all-pass rate, nearly double the next best models, but the best model still fails to fully complete 86% of professional legal deliverables, highlighting that frontier legal work remains largely unsolved.

read4 min views1 publishedJul 14, 2026

All articles Harvey LAB-AA (Legal Agent Benchmark) is our implementation of Harvey's new agentic legal benchmark, evaluating language models on real-world legal work across 24 practice areas.

Models are tested on a private set of 120 legal tasks built by the team at Harvey, spanning practice areas from corporate M&A and capital markets to tax, litigation, and bankruptcy. Models work to create the legal outputs specified in each task, and each task is graded against a rubric of binary criteria. The primary metric we present is the all-pass rate: the share of tasks where all criteria in the rubric are satisfied, reflecting the high standard of real-world professional legal deliverables.

Score #

Harvey LAB-AA: All-pass Rate

Claude Fable 5 (max, with Opus 4.8 fallback) leads Harvey LAB-AA with a 14.2% all-pass rate, after falling back to Claude Opus 4.8 on only one task. This is almost double the next best models, Claude Opus 4.8 (max) and GLM-5.2 (max), which tie at 7.5%, followed by MiniMax-M3 at 6.7% and Claude Sonnet 5 at 5.0%.

Frontier legal work is far from solved: most models pass a majority of individual rubric criteria but fully satisfy the requirements of very few tasks. The best model still leaves ~86% of professional legal deliverables incomplete, and 13 of the 28 models evaluated at launch fully pass zero tasks. Only four models score above 90% on criterion pass rate: Claude Fable 5 (93.6%), Claude Opus 4.8 (91.1%), GLM-5.2 (91.0%), and Claude Sonnet 5 (90.1%).

For up-to-date results see the Harvey LAB-AA evaluation page. Charts show data as at 7 July 2026.

Cost #

Harvey LAB-AA: Cost per Task

Token Usage #

Harvey LAB-AA: Output Tokens per Task

Speed #

Harvey LAB-AA: Time per Task

Turns #

Harvey LAB-AA Benchmark Leaderboard: Average Turns per Task

Score vs. Release Date #

Harvey LAB-AA: All-pass Rate vs. Release Date

Example Tasks & Submissions #

Browse representative Harvey LAB tasks from the public task set, the reference files each model was given, and the deliverables it produced.

Instructions

Review the attached acquisition data room contracts and internal memo for change of control and assignment provisions, and prepare a comprehensive deal team report.

Output: coc-analysis-report.docx Deliverables

Expected outputs the model must produce

  • coc-analysis-report.docxA comprehensive deal team report analyzing change of control and assignment provisions across the target’s material contracts.

Reference files

Provided to the model

Model submissions

Deliverables produced by each model

Open

How Harvey LAB-AA differs from Harvey's LAB #

Harvey LAB-AA is our independent reimplementation of Harvey's evaluation, and there are several key differences to the original version:

  • Models are run on our Stirrupagent harness, enabling features such as context compaction rather than failure when reaching context limits, with simplified Artificial Analysis-authored agent and judge prompts - We do not include Harvey's custom tools and document-generation skill scripts (e.g. pptx, docx), instead providing a simple code execution tool to reflect raw model ability
  • Deliverables must match the exact filename specified, rather than fuzzy matching when models produce incorrect filenames
  • Grading uses a single Gemini 3.1 Pro judge, tested to be well-calibrated against a frontier panel

Harvey LAB-AA resources #

[GitHub](https://github.com/harveyai/harvey-labs) - Harvey LAB-AA runs on
[Stirrup](https://github.com/ArtificialAnalysis/Stirrup), our open-source agent framework

Read the latest

How GPT-5.6 Sol, Terra, Luna compare on intelligence vs cost

GPT-5.6 Sol and Luna are ahead of Terra at every point on the Intelligence vs Cost per Task chart. GPT-5.6 Luna stands out as a particularly cost efficient model

July 13, 2026

Muse Spark 1.1: Meta gains 8 Intelligence Index points in three months

Meta's Muse Spark 1.1 scores 51 on the Artificial Analysis Intelligence Index and is cost and token efficient compared to its peers

July 10, 2026

GPT-5.6 benchmarks across Intelligence, Speed and Cost

GPT-5.6 Sol comes close second to Claude Fable 5 in the Artificial Analysis Intelligence Index at one third of the cost, and leads the Artificial Analysis Coding Agent Index in OpenAI’s Codex harness

July 9, 2026

── more in #artificial-intelligence 4 stories · sorted by recency
── more on @harvey 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/harvey-lab-aa-evalua…] indexed:0 read:4min 2026-07-14 ·