What's working
- Agent Builder turns client workflows into durable switching costs.
- LexisNexis content deal closed the data credibility gap fast.
- Embedded engineers inside accounts compress churn to near zero.
Harvey just closed a $200M round at an $11B valuation and is now planting embedded legal engineering teams inside its largest customers. That is not a product update. That is an account-lock strategy. If you sell AI legal tools to law firms or in-house teams, this profile tells you exactly where Harvey is strongest, where the floor is, and which buyers it structurally cannot serve.
Harvey now hosts 25,000+ custom agents built by clients and is embedding its own legal engineers inside accounts to build more. Every agent a firm creates on Harvey is institutional switching cost that raises the exit price for that client.
GTMThe June 2025 alliance gave Harvey exclusive generative AI access to LexisNexis primary US law, statutes, and Shepard's Citations inside its platform. This closed the one credibility gap competitors could exploit and forces every other vendor to rethink their content strategy or find an acquisition target.
PricingA 20-seat minimum at roughly $1,200 per lawyer per month sets a hard floor of approximately $288,000 annually before add-ons. Harvey's pricing intentionally excludes small firms, solo practitioners, and most mid-market in-house teams. That exclusion is also a market map for every founder in this category.
NarrativeHarvey's homepage and March 2026 fundraise announcement frame the company as the system through which legal work gets done, not a point tool. Combined with Vault, Workflow Builder, and a Microsoft 365 Copilot integration in Q2 2026, the narrative is now backed by enough surface area to be credible to a firm's IT and procurement function, not just the practice group.
GTMHarvey went from a $3B valuation in February 2025 to $11B in March 2026, raising over $1B total. That capital funds embedded engineering teams, M&A (Hexus acquired January 2026, Lume AI talent deal), and international expansion now underway across 60+ countries. For a YC-stage founder, this is relevant because it defines what Harvey can afford to do in the segments it does want.
Not raw changes. Directional evidence across product, pricing, content, and market motion.
We track real changes across pricing, positioning, and product. You get clear signals in one place and push them to your team instantly.
Works with the communication tools you already use
Fast Company
Confirms Harvey's 250+ platform improvements and Workflow Builder adoption as the core of its stickiness narrative, corroborating the lock-in thesis.
Artificial Lawyer
Independent legal tech commentary calling this possibly the most important legal tech move in a decade, validating the data moat signal.
CNBC
Confirms $190M ARR, 100,000+ lawyers across 1,300 organizations, and CEO's explicit agent expansion mandate from the March 2026 round.
Public review summary
Harvey has zero reviews on G2 and Capterra by design: enterprise-only sales with no self-serve means no public review volume. Gartner Peer Insights carries a small number of positive ratings from legal operations professionals, all skewing 4 to 5 stars.

Toarn AI
Public signal synthesis
Grade B · Qualitative signals from Gartner Peer Insights are positive, but near-zero public review volume makes a reliable aggregate grade impossible.
Sources: Gartner Peer Insights
Harvey deliberately avoids G2 and Capterra; the absence of reviews is a sales strategy choice, not a satisfaction signal. Grade should be treated as directional, not definitive.
Leadership signal
Winston Weinberg (CEO) tied the March 2026 $200M raise explicitly to scaling AI agents and growing embedded legal engineering teams globally, a clear signal that headcount investment is shifting from sales to customer success and in-account engineering.
Executive summary · Read this first
Harvey has moved from AI assistant to what it now calls the operating system for legal work. More than 25,000 custom agents run on the platform, the majority of AmLaw 100 firms are signed up, and fresh capital is funding embedded legal engineering teams that sit inside client organizations to build and improve those agents. That is a retention play, not a growth play, and it is the most structurally durable move in the category.
The LexisNexis alliance from June 2025 closed the last credibility gap Harvey had: access to primary law content and Shepard's Citations inside the Harvey interface. Combined with Agent Builder, Vault for up to 100,000 documents, and a Microsoft 365 Copilot integration rolling out in Q2 2026, the platform is pulling every major legal workflow into one renewal line.
Harvey's published floor is approximately $288,000 per year for a 20-seat minimum, enterprise-only, no self-serve. That is both its strongest asset and your clearest opening. Every firm below that threshold, every solo, boutique, or mid-market in-house team, and every workflow Harvey prices too high to touch is a real buyer your company can own.
The window is not closing uniformly. Harvey will deepen in BigLaw and Fortune 500 legal. The question for you is whether you are building something that wins the segments Harvey leaves on the table, or whether you are trying to beat it head-on in a market where it has $1B in capital and embedded engineers in the accounts.
CoCounsel reached 1 million professional users across 107 countries by February 2026, and Thomson Reuters announced next-generation agentic CoCounsel Legal entering beta with conversational task execution grounded in Westlaw and Practical Law content.
Legora competes directly with Harvey on enterprise contract review and M&A diligence workflows at European law firms, positioning on structured document analysis and pricing transparency as differentiators against Harvey's opaque enterprise quotes.
Spellbook reported more than 4,000 in-house legal teams and law firms using its platform for contract drafting and review as of early 2026, with per-seat pricing and self-serve access targeting the mid-market segment Harvey's minimums structurally exclude.
Noise
Product · Q4 2025 to Q2 2026
Infrastructure lock-in over land-and-expandHarvey launched Agent Builder in March 2026, enabling firms to create custom no-code agents on top of the platform. More than 25,000 custom agents now run on Harvey. The $200M raise announced the same month is being used to grow the embedded legal engineering teams that help clients build and maintain those agents inside their own organizations.
Each custom agent a firm builds on Harvey encodes that firm's internal playbooks, style guides, and matter templates. Migrating to a competitor requires rebuilding all of it from scratch. Embedded Harvey engineers accelerate agent creation and create a direct channel into the firm's legal operations budget. This is a fundamentally different retention model than a typical SaaS seat license.
Harvey is replicating the professional services model that kept legacy legal tech vendors entrenched for decades, but at AI speed and with a fraction of the headcount cost. Once a firm has 50 custom agents and a dedicated Harvey engineer, the renewal conversation is not about price. That is a very hard position to displace from the outside.
High impact
Strong: agent count (25,000+), embedded engineering team expansion, and explicit CEO commentary on capital allocation all point the same direction across multiple independent sources.
Build now: define what your own retention model looks like before Harvey's embedded engineers reach your target accounts.
GTM · Q3 2025 to Q1 2026
Data moat formalizedIn June 2025, LexisNexis integrated its primary US law content, statutes, and Shepard's Citations directly into Harvey through the LexisNexis Protege service. Harvey users now access citation-validated legal research without leaving the Harvey interface. Co-developed litigation workflows, including Motion to Dismiss and Motion for Summary Judgment tools, followed immediately.
Before this deal, Harvey's biggest credibility objection from litigators and research-heavy practices was the absence of a verified legal database. That objection is gone. Harvey can now sell into practices that Thomson Reuters' CoCounsel owned by default, and every other vendor without a comparable data partnership faces a structural content gap that is very hard to close quickly.
The alliance is structured so Harvey remains the interface and LexisNexis supplies content, which means Harvey retains the client relationship and ARPU control. It also likely signals RELX is positioning for an eventual acquisition, which would change the competitive map entirely.
High impact
Strong: joint press release from both companies, confirmed integration live in Harvey platform, co-developed workflows publicly announced.
Act now: if your product positions on research quality or citation accuracy as a differentiator against Harvey, that positioning needs to be rebuilt around a specific workflow or buyer segment Harvey cannot credibly cover.
Product · Q1 2026 to Q2 2026
Distribution channel expansionA Microsoft 365 Copilot integration is launching in Q2 2026, allowing lawyers to invoke Harvey directly inside Copilot for agreement analysis, legal research, negotiation support, and precedent retrieval. Harvey already runs on Microsoft Azure and has Outlook and Word add-ins with 12,000+ queries per week in Outlook alone.
Microsoft 365 is already the default workspace for most large law firms and enterprise legal departments. Appearing inside Copilot means Harvey can surface in procurement conversations that start with Microsoft rather than with Harvey's own sales team. It also creates a second distribution channel that does not require Harvey's enterprise sales cycle, which is critical for reaching mid-tier accounts faster.
This is a volume move, not a margin move. Harvey's ARPU from Copilot integrations is likely lower than direct enterprise contracts, but the reach into accounts that would not otherwise take a Harvey sales call is real and compounding.
Medium impact
Moderate: Copilot integration confirmed in published sources for Q2 2026 launch; adoption rate and commercial terms are not yet publicly disclosed.
Monitor this quarter: if your product also lives inside Microsoft 365, Harvey's Copilot presence will compete for the same surface. Define your differentiation inside that environment before it launches.
Ongoing competitor monitoring
Founders and CEOs of AI legal tools companies, particularly YC-stage teams building in the AI Legal Tools category adjacent to or below Harvey's enterprise pricing threshold.
Signal-based, publicly observable claims only. No leaked, private, or proprietary data was used. All valuation and ARR figures are sourced from company press releases or named investor statements.
Sources consulted: Harvey homepage and platform pages, Harvey blog and changelog, Harvey ROI calculator, Wikipedia (Harvey software), Sacra revenue profile, CNBC and TechCrunch funding coverage, LexisNexis and Harvey joint press releases, Fast Company 2026 Most Innovative Companies profile, Gartner Peer Insights review data, third-party pricing analyses (Contracko, CostBench, SoftwareFinder), Artificial Lawyer editorial analysis, and Thomson Reuters CoCounsel public announcements for competitive context. Minimum seven independent surface types consulted. Profile period Q4 2025 to Q2 2026.
This report is compiled from publicly available sources only. No personal information or personal data as defined under applicable privacy laws was collected or processed. All analysis reflects editorial interpretation of public signals, not statements of fact. No guarantee is made as to accuracy, completeness, or timeliness. Business decisions based on this report are solely the reader's responsibility. Toarn accepts no liability for outcomes resulting from reliance on this analysis. Not affiliated with Harvey AI.
Q2 2026 · Updated Apr 11, 2026