Harvey
AI trained for large law firms and enterprise legal teams
An AI platform for legal research, document analysis and drafting built specifically for elite law firms and in-house legal departments, deployed as custom workflows on a firm's own document set.
Picked for AI Tools for Regulated Industries.
Harvey builds AI tools for legal research, document review and drafting, aimed squarely at the top of the market: elite law firms and enterprise in-house legal departments rather than solo practitioners or small firms. Its core differentiator is Vault, a permissioned repository that lets a firm ground every AI answer in its own precedents and playbooks rather than generic legal training data, which matters a lot in a field where firm-specific style and prior work product carry real weight.
It’s built for associates and in-house counsel doing research, memo drafting and contract redlining at volume, with Word, Outlook and Microsoft 365 Copilot integrations that keep the work inside documents lawyers are already editing rather than a separate app. The company has raised at a scale that’s pushed rapid feature releases and made it one of the most closely watched products in legal tech.
The clear limitation is price: enterprise contracts run from tens of thousands of dollars a year at the low end into six figures for larger deployments, with reported double-digit percentage renewal increases on top. That puts it out of reach for smaller firms regardless of how good the product is, and like every legal AI tool, its output still needs review by a lawyer who understands the matter — it’s a research and drafting accelerant, not a decision-maker.
Features
Vault
A curated, permissioned repository of a firm's own precedents, templates and playbooks that grounds every answer in its actual work product.
Assistant
Answers legal questions, drafts memos and analyzes uploaded contracts with citations back to source documents rather than an opaque summary.
Word and Outlook add-ins
Runs legal research and contract redlining directly inside Word and Outlook, so lawyers don't have to leave the documents they're already working in.
Knowledge sources
Queries run against curated legal databases and the firm's own precedent together, so an answer cites something checkable.
Agentic workflows
Multi-step tasks such as diligence review or deposition prep run as a defined sequence rather than as a chain of hand-written prompts.
Multi-document review
Reads across a whole data room at once, surfacing the clauses and inconsistencies that a linear read finds on day four.
Use cases
- Running legal research and first-pass memo drafts for associates
- Redlining commercial contracts against a firm's standard playbooks
- Giving in-house counsel a searchable assistant over the company's own legal precedent
- Running diligence across a data room of thousands of documents
- Summarising a filing set for a client update without a junior reading all of it
- Finding the firm's own precedent for a clause instead of drafting it again
Compare Harvey head to head
One side-by-side comparison, on pricing, platforms and where each one wins.