AI built to the standards of the world's leading law firms
HQ: San Francisco, CA, United States (offices in New York, Toronto, Bengaluru, Dublin, Paris, Singapore and Chicago) | Founded: 2022 | Employees: Several hundred (not disclosed; about 10% of the team are former lawyers in customer-success roles) | Stage: Late-stage private — $550M round at a $15.5B valuation (September 2026); about $1.78B raised in total | Website: https://www.harvey.ai
Harvey makes AI software for legal work. Lawyers use it to research case law, analyse and summarise contracts, run due diligence, and draft documents, with answers grounded in sources they can check. It is sold to large law firms and in-house legal departments, and the customer list includes most of the biggest firms in the world.
The company was founded in 2022 by Winston Weinberg, a former securities and antitrust litigator at O'Melveny & Myers, and Gabriel Pereyra, previously a research scientist at DeepMind and Meta AI. Their insight was that legal work is high-stakes and document-heavy — exactly the kind of work where a careful AI assistant is valuable and where firms will pay a premium for enterprise-grade security.
Harvey's product has evolved from a single assistant into a platform. There are hundreds of pre-built agents for common tasks, a no-code Agent Builder for firms to make their own, a model selector that routes work to different frontier models, and integrations across the legal stack. A deep alliance with LexisNexis embeds authoritative US case law and Shepard's citations directly into Harvey's answers, while partnerships with Microsoft, DocuSign and Intapp extend it into the tools lawyers already use.
The company now sells beyond law firms: insurance, financial services and other professional-services teams with the same document and compliance problems. That expansion is the core of its long-term story, and the reason investors have been willing to price it at a premium to almost every other vertical software business.
Harvey is the clearest winner of the legal AI wave, and the reason is as much distribution as technology. It started in 2022 as a legal copilot for elite law firms and now sits inside 80% of the Am Law 100, five of the Fortune 10, and 1,500+ customers across 60+ countries. Revenue went from about $100M ARR in August 2025 to roughly $400M by August 2026 — a pace almost unmatched outside of foundation labs.
The strategy shifted in an important way. Harvey initially tried to train its own specialised legal model, then scrapped that when frontier models from OpenAI, Anthropic, Google and xAI beat it on its own legal benchmark. Instead of competing on the model, Harvey now wraps frontier models in pre-configured legal workflows, a model-selector and deep integrations into the tools lawyers already use — LexisNexis case law, Microsoft 365 Copilot, DocuSign and Intapp. That is a defendable position because it is about workflow and trust, not raw intelligence.
The funding history reads like a fever chart: $100M at a $1.5B valuation in July 2024, then rounds in February, June, October and December 2025, then $11B in March 2026, then $15.5B in September 2026. The risk in that trajectory is real. Legal reasoning is being commoditised by the same foundation labs Harvey depends on, well-funded rival Legora is close behind, and the price of seats is premium enough that customers will scrutinise the return. Harvey's answer is to move from a legal tool into a broader platform for professional services — the same play that worked for it so far, one document type at a time.
Browse: Home | Blog | About | All Startups A-Z | View full profile on StartupWiki