Thomson Reuters CoCounsel

Published · Researched 2026-09-22

Thomson Reuters CoCounsel — AI legal research

Maker: Thomson Reuters (Westlaw's parent). CoCounsel was built by Casetext — founded 2013 by Jake Heller, Pablo Arredondo, and Laura Safdie — which pivoted its entire ~120-person team to build CoCounsel on GPT-4 within days of its release, layered over Casetext's legal databases and prompt engineering. Thomson Reuters acquired Casetext for $650M in 2023. (Company history per Frank Ramos's LinkedIn comparison post, ~June 2026.) Purpose: AI legal assistant for research and litigation support across practice areas — "skills" packaged as discrete workflows (deposition prep, contract analysis, database search) rather than an open chat, which practitioners say reduces hallucination risk in practice. Current 2026 version: CoCounsel (Core and All-In tiers) deeply integrated with the Thomson Reuters ecosystem — Westlaw, Practical Law — plus a Microsoft Word add-in. Research sibling: Westlaw Precision with CoCounsel / Westlaw AI-Assisted Research.

Social standing

Platform pulse:

• LinkedIn: Frank Ramos ("Miami Mentor," ~June 2026, post on "Westlaw's CoCounsel vs. Lexis+ AI with Protégé") gave a detailed take: "Both are strong platforms. Both do excellent work. But if I'm choosing, I'm going with Lexis." His key points: CoCounsel was "never proprietary AI built in-house — it was a GPT-4 wrapper layered on Casetext's legal databases and prompt engineering. That's not a knock; it was brilliantly executed"; the TR acquisition means "it's never truly yours. Integrating an outside platform into your own stack is harder than building from scratch"; and price — "Westlaw Precision with CoCounsel (which runs around $428/user/month for solos in many markets)" vs. Lexis+ AI generally more competitive. He also noted the Harvey–LexisNexis alliance "cuts against Westlaw."
• Threads: @sung.kim.mw (2026-08-28, 9 likes / 3 replies) on Thomson Reuters' open model Thomson-1.0-Small, "post-trained on Thomson Reuters' proprietary data — decades of news, contracts, regulatory filings, case law, statutes and practitioner guidance" (huggingface.co/thomsonreuters/Thomson-1.0-Small) — framed as a possible substitute for subscribing to TR services.
• Web community: scored.tools (2026) rates CoCounsel 8.9/10 — "the workhorse research tool that keeps showing up in firm stacks because it just works… Its tight integration with Westlaw's underlying database is the moat." toolify.ai's community review roundup: "Users have generally praised CoCounsel for its efficiency, accuracy, and ease of use… Some have noted occasional inaccuracies or limitations in its knowledge base." growlaw.co's 2026 legal-research ranking gives Casetext (CoCounsel) 5/5: "Deep Westlaw integration, agentic AI, real-time document analysis… ~$225/user/month."

Recurring praise: just works inside the Westlaw ecosystem; discrete skill-based workflows reduce hallucination risk vs. open chat; strong for litigation-focused research and document analysis. Recurring complaints: AI price is not the bill — Westlaw subscription usually required underneath; acquisition-integration seams (Casetext product inside TR stack); the most serious black mark: United States v. Farris (6th Cir., April 2026) — a brief drafted with CoCounsel's help reached the court with fabricated quotations attributed to real cases; the attorney was denied court-appointed compensation, referred for discipline, and removed from the case (reported by Charles Stack, Medium, Aug 2026 — the tool wasn't the villain; insufficient human review was).

Verified quotes (platform + context):

• "CoCounsel was never proprietary AI built in-house — it was a GPT-4 wrapper layered on Casetext's legal databases and prompt engineering. That's not a knock; it was brilliantly executed." — Frank Ramos, LinkedIn (~June 2026).
• "Lexis+ AI is generally more competitively priced than Westlaw Precision with CoCounsel (which runs around $428/user/month for solos in many markets)." — Frank Ramos, LinkedIn (~June 2026).
• "Cocounsel (Thomson Reuters) is the workhorse research tool that keeps showing up in firm stacks because it just works." — scored.tools (2026).

Social validation: Tier 2

Two independent social sources with identifiable users: LinkedIn (Frank Ramos detailed comparison, practitioner-authored) and Threads (@sung.kim.mw open-model post with reply discussion). Corroborated by web community reviews (scored.tools, growlaw, toolify.ai). Facebook Groups: no CoCounsel discussion found (Sept 2026). Instagram: not surfaced.

Feature table (2026)

Pricing (third-party reported, UNVERIFIED by vendor)

• Core ~$225/user/month; All-In ~$850; Westlaw research subscription often required underneath at ~$200–$400 (Codebridge, checked 21.09.2026).
• ~$428/user/month for solos in many markets (Frank Ramos, LinkedIn, ~June 2026).
• Quote-only; no published pricing.

Integrations

Westlaw, Practical Law, Microsoft Word add-in; TR ecosystem (Document Intelligence lineage).

Support reputation

Enterprise vendor support via Thomson Reuters; no verifiable public support-rating data found — UNVERIFIED.

Community verdict

CoCounsel is the incumbent's AI that practitioners actually keep in their stack — less flashy than Harvey, more embedded where research already happens. The community's honest accounting: it works, the Westlaw backbone is the reason, but the total bill (AI tier + underlying research subscription) is the real number to negotiate, and the Farris sanctions episode is a permanent reminder that its citations must be pulled and read by a human before filing.