Home › Best AI by category › Contract review › Report
As of Oct 26, 2026, aiforlawyers.app counts 305 lawyer voices on contract review; 248 voices name one of 10 tools, 4 of which have 10+ voices and are rated. Highest floor of the 90% band: Claude (net +46, band +29 to +59, 65 voices, Medium confidence).
Download the PDF (cutoff Oct 26, 2026)Data (JSON)Ranking page
Cite this
aiforlawyers.app (AI Vortex), Contract review report, cutoff Oct 26, 2026, https://aiforlawyers.app/report/contract-review, licensed CC BY 4.0.
For one figure: “As of Oct 26, 2026, aiforlawyers.app counts 305 lawyer voices on contract review; 248 voices name one of 10 tools, 4 of which have 10+ voices and are rated. Highest floor of the 90% band: Claude (net +46, band +29 to +59, 65 voices, Medium confidence).”
10 tools named in 305 voices on contract review. Rated tools first, ordered by the floor of the 90% band; then the tools with few voices, with their count and no band. Net = like minus dislike per 100 voices.
| # | Tool | Status | Voices | Net | 90% band | Confidence | Top complaint (voices) |
|---|---|---|---|---|---|---|---|
| 1 | Claude | Rated | 65 | +46 | +29 to +59 | Medium | Accuracy (12) |
| 2 |  | Rated | 51 | +39 | +20 to +55 | Low | Accuracy (9) |
| 3 |  Harvey | Rated | 43 | +12 | −7 to +30 | Low | Accuracy (17) |
| 4 |  ChatGPT | Rated | 57 | −11 | −29 to +9 | Low | Accuracy (22) |
| – |  Copilot | Few voices | 9 | – | no band: 9 of 10 voices needed | not rated | Accuracy (3) |
| – | CoCounsel | Few voices | 6 | – | no band: 6 of 10 voices needed | not rated | Accuracy (3) |
| – |  Legora | Few voices | 6 | – | no band: 6 of 10 voices needed | not rated | Accuracy (2) |
| – |  Luminance | Few voices | 6 | – | no band: 6 of 10 voices needed | not rated | Accuracy (4) |
| – | Gemini | Few voices | 3 | – | no band: 3 of 10 voices needed | not rated | none on record |
| – | Perplexity | Few voices | 2 | – | no band: 2 of 10 voices needed | not rated | none on record |
The most named complaints per tool, with a verbatim quote each. Counts are voices, not opinions of the firm or of us.
18 voices of 65 criticize it or are mixed about it with a named complaint: accuracy (12), privacy (2), integration (2), hallucination (1), trust, liability (1).
“Explaining to people why Claude’s positions are unreasonable has become a part of my day to day job.”
“Howver using claude for building own contract analyzer , won't give you data privacy as you would be using claude anyway.”
Rated · 51 voices19 voices of 51 criticize it or are mixed about it with a named complaint: accuracy (9), price (5), hallucination (3), trust, liability (1), privacy (1).
“Out of say, 30 comments it had in analyzing an agreement, most seemed like noise (possibly technically valid, but not helpful) and one or two were interesting and useful but far from being worth the fee.”
“I don’t mind useful wraps, but the price point in this one is so steep…”
HarveyRated · 43 voices24 voices of 43 criticize it or are mixed about it with a named complaint: accuracy (17), hallucination (2), trust, liability (2), price (2), integration (1).
“I use Harvey, Claude, etc. for diligence and know for a fact it is still leagues behind in picking up change of control/anti-assignment provisions.”
“I wish I could upload my back and forth with Harvey today where it incorrectly cited sections over and over again and made up signatories in a checklist.”
ChatGPTRated · 57 voices32 voices of 57 criticize it or are mixed about it with a named complaint: accuracy (22), hallucination (5), trust, liability (2), privacy (2), integration (1).
“I find that when I feed contracts to ChatGPT (which I do all the time) it will try to find 50 things wrong with it.”
“I uploaded a contract and asked it a question and it started citing nonexistent sections.”
CopilotFew voices · 9 of 103 voices of 9 criticize it or are mixed about it with a named complaint: accuracy (3).
“Even procurement is running contracts, order forms etc through whatever GenAI just to produce completely unusable risk lists.”
4 voices of 6 criticize it or are mixed about it with a named complaint: accuracy (3), trust, liability (1).
“My CoCounsel demo was a disaster. Spellbook and its ilk are good.”
“I've used AI stuff like Spellbook, CoCounsel, AI Lawyer, and the pattern I keep seeing is that the analysis label doesn't matter if the UI encourages people to accept changes like autopilot.”
LegoraFew voices · 6 of 104 voices of 6 criticize it or are mixed about it with a named complaint: accuracy (2), hallucination (1), price (1).
“Legora and Harvey price most law firms and even some in house teams out, but in my trials, they don’t add any value on the contract analysis side and the results of the suggested redlines are better with other tools.”
“We also tried Legora's tabular review, there were also numerous halucinated provisions in our due diligence exercise.”
LuminanceFew voices · 6 of 104 voices of 6 criticize it or are mixed about it with a named complaint: accuracy (4).
“My comments about how Luminance didn't work at all for me have been downvoted multiple times and there's a four year old account with zero previous comments coming to life and saying Spellbook is great.”
No named complaint on record among its 3 voices.
No named complaint on record among its 2 voices.
Public tests that include contract review work, with their own publisher, date and caveat. They measure question sets, not lawyers’ experience, and are never added to the ranking above. Others are on the arena.
legal-specific vendor test · Vals AI
published 2025-02-27 · read 2026-10-22
Four tools were tested on seven tasks against a lawyer baseline: CoCounsel, Vincent AI, Harvey Assistant and Oliver (Lexis+ AI withdrew from the sections studied). Harvey took the top score on five of the six tasks it entered, including Document Q&A at 94.8% and Chronology Generation at 80.2%; CoCounsel scored 89.6% on Document Q&A and 77.2% on Document Summarization, the top score there. The lawyer baseline beat the AI tools on EDGAR research and on Redlining (79.7%).
Caveat. Vendors opted in task by task (Harvey six of seven, CoCounsel four), so tools were not compared on the same tasks; 2025 versions, over a year old; not a ranking of all legal AI.
Names: Harvey Assistant, CoCounsel, Vincent AI (vLex), Oliver (Vecflow, now August), Lexis+ AI (withdrew).
legal-specific vendor test · Harvey (benchmark) / Mercor (leaderboard)
published 2026-05-06 (launch per Harvey blog coverage); leaderboard date not shown on page · read 2026-10-22
Open-source benchmark of 1,749 tasks across 24 practice areas plus in-house contracting, graded with all-pass rubrics (100,000+ criteria). Pass@1 top 5 on the page: Gemini 4 Argon High 27.4% (±5.4), Grok 4.7 xHigh 19.5% (±4.5), Muse Spark 1.2 xHigh 19.4% (±4.6), Claude Opus 5 Max 17.8% (±1.5), Claude Fable 5.1 Max 17.4% (±4.2).
Caveat. Built by Harvey (a vendor); the leaderboard lists foundation models, not Harvey's own product; leaderboard publication date not shown and the top-5 error bars overlap heavily; the leaderboard page counts 1,749 tasks versus 'more than 1,200' reported at launch, so the set has grown.
Names: Gemini 4 Argon, Grok 4.7, Muse Spark 1.2, Claude Opus 5, Claude Fable 5.1.
https://www.mercor.com/apex/partner-benchmarks/harvey-lab-leaderboard/
Related on AI Vortex
Use these numbers: the counts, rates and bands on this report are free to reuse under CC BY 4.0. Credit “aiforlawyers.app (AI Vortex)” with the date and a link to the page. Quotes belong to the people who wrote them: quote them with the link to the original post. Machine-readable: index.json. How to cite.
