FAQs
💬
What is CoCounsel, and who is it built for?
CoCounsel is an AI-powered assistant that handles research, document analysis, and drafting for professionals, mainly lawyers and tax experts. It uses agentic and generative AI to automate tasks like legal memos or tax reviews, grounded in trusted sources like Westlaw and Checkpoint. It’s ideal for firms wanting to cut down on routine work while keeping high accuracy, with over 20,000 law firms already using it, including most Am Law 100 ones.
How does CoCounsel integrate with tools like Westlaw or Microsoft 365?
It connects seamlessly with Westlaw for research, Practical Law for drafting playbooks, Checkpoint for tax insights, and Microsoft 365 apps like Word and Teams for direct edits or queries. This lets you pull context from your daily tools without switching apps, saving time on things like contract reviews in Word. Users report it feels like a natural extension of their workflow.
What are the main features of CoCounsel for legal work?
Key skills include Deep Research for multi-step legal plans, document comparison for spotting differences in contracts, deposition prep by analyzing transcripts, and drafting with built-in citations. It excels at summarizing long docs or building chronologies, often finishing in minutes what takes hours manually. Recent betas add agentic workflows for end-to-end tasks like complaint drafting.
How does the tax version of CoCounsel work, and does it save time?
CoCounsel Tax automates research memos from IRS codes and Checkpoint, reviews returns for issues, drafts letters like penalty abatements, and spots savings opportunities like cost-segregation studies. A tax partner from a top U.S. firm said it handles 80% of a one-hour process, leaving room for judgment. It integrates with Excel and DMS for smooth tax planning.
What about pricing—how much does CoCounsel cost?
CoCounsel Core starts at $225 per user per month, with volume discounts for firms; full access via Westlaw Precision bundles it cheaper for subscribers. Tax versions vary, but expect $400+ monthly for advanced features, billed annually. Trials last 2-3 months, though some users note the "free" part ties into longer contracts. Smaller firms get Essentials plans scaled to growth.
Is CoCounsel secure for handling sensitive client data?
Yes, it uses TLS 1.2 for transit, AES-256 encryption at rest, and zero-retention API calls so prompts aren’t stored or used for training. Data stays in the U.S. (with options in UK, Australia, Canada), and partners like OpenAI can’t access it. Retention follows your policy, and you can delete anytime. Guardrails limit outputs to verified content to cut bias.
How accurate is CoCounsel compared to other AI tools like Harvey?
In 2025 benchmarks, CoCounsel scored high on document Q&A (89.6%) and summarization (77.2%), beating human baselines by over 10 points in some tasks, though Harvey led overall. It’s grounded in Westlaw’s authoritative content to minimize hallucinations, but users still verify outputs. I think it shines for research-heavy work, while Harvey might edge out on custom models.
What do users say about CoCounsel’s pros and cons?
Pros: Speeds up tasks (e.g., contract diffs in minutes), catches errors in briefs, and integrates well; Reddit users call it a game-changer for solos. Cons: Pricey at $679/month for some quotes, occasional slowdowns post-acquisition, and needs verification to avoid rare hallucinations. Overall, it boosts efficiency by 10x for many, but test a trial first.
Can small firms or solos afford and use CoCounsel effectively?
Absolutely, with Essentials plans starting lower and scaling up—no need for Westlaw to start, though bundling saves money. It handles 500-page contract reviews or brief proofreading solo-style, and a 2-month trial lets you test without commitment. Users in small practices praise its document focus over basic research.
What’s new with CoCounsel in late 2025, like agentic AI?
Recent betas include agentic workflows for delegating full tasks (e.g., bulk review of 10,000 docs), customizable plans, and Deep Research on Practical Law. Rollouts continue into 2026, aiming for custom agents. It’s shifting from prompts to execution, with users seeing 63-80% time cuts on reviews.