CoCounsel is Thomson Reuters' AI assistant, built on the Casetext platform they acquired for $650 million in 2023. It's designed to layer AI on top of Westlaw's legal research database. The one thing attorneys need to know: CoCounsel's biggest strength — Westlaw integration — is also its trap, because it locks you deeper into the Thomson Reuters ecosystem.
What CoCounsel (Thomson Reuters) Actually Does
CoCounsel handles legal research with real Westlaw citations, document review, contract analysis, deposition preparation, and timeline creation from document sets. The Westlaw connection is the key differentiator — when CoCounsel cites a case, it's pulling from the actual Westlaw database, not generating citations from training data. That dramatically reduces hallucination risk on case citations.
In daily practice, attorneys use CoCounsel most for research queries where they need verified authorities fast. Instead of running Boolean searches in Westlaw, you describe what you need in plain language and CoCounsel returns relevant cases with proper citations. It also handles document review tasks — upload a contract and ask it to identify specific provisions, risks, or deviations from standard terms.
The deposition prep and timeline features are newer and less proven. They work best with structured document sets (chronological emails, medical records, transaction documents) and struggle with messy, inconsistent document productions. Attorneys report the timeline feature saves time on simple chronologies but still requires heavy editing for anything complex.
Pricing and Lock-In
CoCounsel is sold as an add-on to existing Westlaw subscriptions or bundled with new Westlaw contracts. Estimated pricing is $100-200 per seat per month on top of existing Westlaw fees. For a firm already paying $200-400/user/month for Westlaw, CoCounsel pushes the total Thomson Reuters bill to $300-600/user/month.
That's the real cost calculation most firms miss. CoCounsel isn't a $100-200/month tool — it's a $300-600/month ecosystem commitment. And it's not available standalone. You can't buy CoCounsel without Westlaw. Thomson Reuters designed it that way deliberately: CoCounsel makes Westlaw stickier, and Westlaw makes CoCounsel possible.
Compare this to running Claude ($25/user/month) with manual citation verification on Google Scholar or a Fastcase subscription ($95/month). You lose the seamless citation integration, but you save $200-500/user/month. For a 20-attorney firm, that's $48,000-$120,000 in annual savings. The question is whether citation convenience justifies that spread.
Best Use Cases
CoCounsel makes the most sense for firms already deep in the Westlaw ecosystem. If you're paying for Westlaw, Practical Law, and other Thomson Reuters products, CoCounsel is a natural extension that adds genuine value to your existing spend. The AI layer makes Westlaw faster to use, which means your research subscription generates more output per dollar.
Litigation firms doing heavy case law research benefit most. CoCounsel's ability to find relevant authorities from natural language queries, with verified citations, saves 30-60 minutes per research task compared to manual Westlaw searches. For firms billing 20+ research hours per week, that adds up.
The document review features work well for transactional practices handling contract review at scale. Upload a set of contracts, ask CoCounsel to compare terms, flag deviations, or extract specific provisions. It's faster than manual review and the Westlaw backing gives confidence in the legal analysis.
Limitations and Honest Take
CoCounsel creates double lock-in. You need Westlaw to use it, and once you're using CoCounsel, leaving Westlaw means losing both your research platform and your AI assistant simultaneously. That's not an accident — it's the business model. Thomson Reuters spent $650M on Casetext specifically to make their ecosystem harder to leave.
The tool is narrowly focused on what Westlaw data enables. It's strong on case law research and legal analysis. It's weaker on general drafting, client communications, business strategy, and anything that doesn't map directly to Westlaw's database. For the 50% of legal work that isn't research — emails, memos, client updates, billing — CoCounsel doesn't help.
Pricing transparency is poor. Thomson Reuters negotiates individually with each firm, which means smaller firms pay more per seat than large firms with negotiating leverage. There's no published price list, no self-service signup, and no way to trial CoCounsel without engaging a Thomson Reuters sales team. That friction is by design.
When to Use CoCounsel (Thomson Reuters) vs Building Your Own
Use CoCounsel when your firm is already on Westlaw with no plans to leave, does heavy litigation research, and values verified citations enough to pay the premium. In that scenario, CoCounsel is an efficiency multiplier on money you're already spending. The incremental cost ($100-200/seat/month) is justified if each attorney saves 3-5 hours of research time per month.
Build your own workflow when you're not locked into Westlaw, when your practice doesn't revolve around case law research, or when you're cost-sensitive. Claude or ChatGPT at $20-25/month, combined with free citation verification through Google Scholar, CourtListener, or Fastcase, covers 80% of CoCounsel's research capabilities. You spend an extra 10-15 minutes per query verifying citations manually, but you save $100-200/seat/month.
The breakeven is clear: if CoCounsel saves each attorney more than 2 hours of research per month (at the firm's internal cost rate), the $100-200 add-on pays for itself. If adoption is spotty or your practice isn't research-heavy, you're paying for convenience you don't use enough to justify.
The Bottom Line
CoCounsel is a strong research tool trapped inside an expensive ecosystem. If you're already paying for Westlaw and do heavy litigation research, it's worth the add-on. If you're not on Westlaw, don't let CoCounsel be the reason you sign that contract — the total cost of the ecosystem far exceeds the AI's value.
AI-Assisted Research. This piece was researched and written with AI assistance, reviewed and edited by Manu Ayala. For deeper takes and the perspective behind the research, follow me on LinkedIn or email me directly.