Harvey
Also known as: Harvey AI Corporation
Enterprise legal AI platform with purpose built and custom agents, Vault document analysis, research, memory and an agent builder, used across most of the Am Law 100.
Harvey is an enterprise legal AI platform for law firms and in house legal teams, sold by Harvey AI Corporation of San Francisco and used by the majority of the Am Law 100. Its agents take a legal task from description to a cited, review ready deliverable: the user previews and approves the plan before work begins, Harvey coordinates multiple agents in parallel, brings the user back in when judgment is needed, and logs every step with a citation on every claim.
Vault stores and bulk analyses the firm's documents, Knowledge researches legal, regulatory and tax questions across licensed sources, Contract Intelligence handles review and negotiation, and Agent Builder and Workflow Builder let firms encode their own playbooks into custom agents, alongside more than five hundred prebuilt agents across practice areas. Harvey Memory carries each user's preferences across sessions and surfaces, with users reviewing or deleting memories and admins granting the feature by role, and Scheduled Agents run reports and reviews unattended.
Harvey works inside Microsoft Word, Outlook and Microsoft 365 Copilot, on mobile, and through connectors to iManage, NetDocuments, Box, SharePoint, Google Drive and Gmail; a documented API and a hosted MCP server let Claude, Gemini and Copilot call Harvey directly. Workspace admins choose which of OpenAI, Anthropic and Google models to enable and users can switch models in Assistant, Vault and Workflow Builder. Data stays in the customer's chosen region (US, EU or Switzerland, or Australia) with SAML SSO, audit logs, IP allow-listing, customer set retention and ethical wall enforcement, under SOC 2 Type II, ISO 27001, ISO 27701 and ISO 42001. Pricing is by demo.
Vendor details
Canonical URL
https://harvey.ai
Category
Enterprise operations agent
Subcategory
Legal — enterprise legal AI platform
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
In practice
Due diligence means a junior associate reading thousands of contracts by hand. Harvey's Vault holds the document set and its agents review across it, so the first pass isn't all human hours.
Your firm has a repeatable workflow you wish ran itself. Harvey's Agent Builder lets teams create custom agents for it, alongside specialized agents for immigration, tax, and M&A work.
A complex matter needs many steps, not a single answer. Harvey's Long-Horizon Agents carry a task across research, drafting, and review rather than stopping at one prompt.
Agentic Index coverage score
12.5 / 14 capabilities · 89%
| Integrations & Tool Calling | Full |
|---|---|
|
Connectors give agents secure access to Microsoft, DMS systems and the other tools the team uses, so they can find information and complete work across systems, and Harvey's own iManage integration documents importing from and exporting work product back to the iManage workspace in one click. The June 2026 Connector Library adds native Google Drive and Gmail integrations and MCP connections to iManage, NetDocuments, PitchBook, Intralinks, Datasite and Box, in Early Access at announcement. SourceHarvey, harvey.ai/platform/agents, harvey.ai/blog/harveys-imanage-integration and harvey.ai/blog/connector-libraryread 2026-09-06 |
|
| Workflow Orchestration | Full |
|
Harvey coordinates a team of agents that split large tasks, work in parallel and return review ready outputs, plans each task before running it, and executes legal work end to end from description to deliverable; Agent Builder and Workflow Builder let customers define their own multi step agents. SourceHarvey, harvey.ai/platform/agentsread 2026-09-06 |
|
| Knowledge Grounding & RAG | Full |
|
Vault stores, organizes and analyzes the customer's legal documents in bulk as persistent projects that are queried with source citations, and Knowledge grounds research in licensed sources such as UK tax law; the developer docs document Vault projects and knowledge sources as addressable resources. Saved preferences live separately, in Memory. SourceHarvey, harvey.ai/platform/agents and developers.harvey.ai/guides/harvey_mcpread 2026-09-06 |
|
| Human Oversight & Guardrails | Full |
|
The Agents page documents a plan gate that is Harvey's own mechanism: the user previews the plan, adjusts the scope and approves work before Harvey begins, and Nudges bring the user in mid run when judgment is needed before work continues. Output review lands in the user's Word document, which is a shared surface, but the approval before the run and the escalation during it are the vendor's. SourceHarvey, harvey.ai/platform/agentsread 2026-09-06 |
|
| Security, Identity & Governance | Full |
|
Certifications listed on the security page include SOC 2 Type II and ISO 27001, along with ISO 27701 and ISO 42001, with a public trust center at trust.harvey.ai. Customers get SAML SSO and IP allow listing, retention policies they set with deletion at any time, and enforcement of the firm's own ethical walls kept in sync. The developer docs add per user permissions and feature gates the workspace admin manages. SourceHarvey, harvey.ai/security and developers.harvey.airead 2026-09-06 |
|
| Observability & Auditability | Full |
|
Every claim in agent output is backed by a citation and every step is logged for verification before partner or client review. The security page lists audit logs customers can use among its standard controls, and an audit log endpoint on the EU API is documented for developers. SourceHarvey, harvey.ai/platform/agents and harvey.ai/securityread 2026-09-06 |
|
| Memory & State Persistence | Full |
|
Harvey Memory is a documented memory layer with a stated scope and lifetime: preferences saved in the flow of work are personal to the user and persist across sessions and surfaces, from the web app and Outlook to the Word add in and playbook runs. They are reviewed, edited or deleted from one place and granted by role through admin controls. Agent Builder configuration is a separate asset. SourceHarvey, harvey.ai/platform/memoryread 2026-09-06 |
|
| Deployment & Data Residency | Full |
|
The customer keeps everything in region, choosing the EU or Switzerland, the US, or Australia, according to the security page, and the login offers separate US, EU and AU application hosts, with an EU hosted API endpoint documented for developers. It is hosted only; no VPC or on premises option is documented. SourceHarvey, harvey.ai/securityread 2026-09-06 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
Harvey ships prebuilt use case agents a customer deploys as they are. The Agents page shows regulatory briefing, deposition preparation, deal point analysis, precedent drafting and case strategy agents as examples, and in May 2026 the vendor announced more than 500 live agents across every major practice area, plus agents developed with A&O Shearman. Remove one and the others still work whole. Agent Builder is the customer's own tooling for building more. SourceHarvey, harvey.ai/platform/agents and the vendor's May 2026 agents announcementread 2026-09-06 |
|
| Triggers & Channel Coverage | Full |
|
Scheduled Agents run reports, updates and reviews automatically on a schedule the customer chooses, so work starts without a person. Users reach agents from the web app, the Word add in and Outlook, from Harvey Mobile, and from outside assistants through the Harvey MCP server. Webhook or event subscriptions are not documented. SourceHarvey, harvey.ai/platform/agents and developers.harvey.airead 2026-09-06 |
|
| Model Flexibility & Routing | Full |
|
A help center release note documents model choice for customers: workspace admins roll out Anthropic Claude and Google Gemini alongside OpenAI models through self serve permissions, and an opt in Model Selector lets users switch models in Assistant, Vault and Workflow Builder, with Harvey routing automatically by default. SourceHarvey, help.harvey.ai release notes (30 June 2025) and harvey.ai/blog/why-harvey-is-multi-model-by-design; eu.help.harvey.ai/release-notes/multi-model-support-now-availableread 2026-09-06 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
developers.harvey.ai documents the Harvey API: an Open Ended completion endpoint on Harvey's own models with bearer token authentication and an EU hosted API host, sold as an extra through the account manager. Alongside it, a remote Harvey MCP server over Streamable HTTP with OAuth exposes ask, Vault query and knowledge source tools to Claude, Gemini and Microsoft 365 Copilot, scoped to each user's permissions and switched on per user by the admin. SourceHarvey, developers.harvey.ai/guides/harvey_mcp and developers.harvey.ai/reference/completionread 2026-09-06 |
|
| Testing, Debugging & Optimization | Partial |
|
Command Center gives administrators analytics, benchmarking and agentic insights on the organization's use of Harvey, and the vendor publishes its own BigLaw Bench model evaluations. No test harness, test cases or release gate for customers is documented for agents built in Agent Builder or Workflow Builder, and the vendor's own model evaluations are its release practice, not something the customer runs. SourceHarvey, harvey.ai/platform/agents and harvey.ai/platform/command-centerread 2026-09-06 |
|
| Browser & Computer Use | Not documented |
|
Agents reach systems through documented connectors, MCP and the API, and Harvey runs as a web app, a Word add in, an Outlook integration and a mobile app, which are the product's own surfaces; no browser, desktop or computer control is documented. SourceHarvey, harvey.ai/platform/agentsread 2026-09-06 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Recent platform changes
A new Connector Library in Harvey is a catalog of admin enabled connectors that are either native integrations or vendor run MCP servers its agents call as tools. Harvey says each MCP tool is security reviewed first, every call passes a policy engine that fails closed, and admins can switch off a whole connector or a single tool.
Bears on: Security / enterprise
View sourceHarvey added agentic search for connected SharePoint, OneDrive, and Google Drive content. Users can have Harvey find relevant documents directly instead of locating and selecting the files themselves.
Bears on: Agent capability
View sourceHarvey lets users hand any Review Table action to the Harvey Agent in plain language, including reclassifying a column, importing metadata as new columns, bulk editing cells and routing flagged rows to a reviewer.
Bears on: Agent capability
View sourcePricing
Contact for pricing
What is public
Harvey publishes no pricing page and no rate card; the site offers Request a Demo only. The Open Ended API is documented as an additional purchase arranged through the account manager.
Billing mechanics
Not disclosed; enterprise contracts quoted through sales.
Cost watchouts
The Open Ended API is an additional purchase through the account manager; no other commercial terms are published.
Additional watchouts
No public pricing, no free tier and no self serve path; the API is an additional purchase.
Sales call required
Yes, required for paid access
Free / trial
No trial or free tier published; request a demo
Lowest paid plan
Not published
Key ambiguities
No official price exists on any first party page.
Support SLA / resale
Enterprise onboarding, dedicated solutions/onboarding teams, and enterprise security/compliance are standard; specifics are contract-defined
Missing data
No pricing page; seat price, seat minimums, term and onboarding fees are not published. Earlier per seat figures came from third party reports and are withdrawn.
Related vendors
- 11th Estate — Agentic platform whose AI engine scans markets, matches an…
- Adopt AI — AI CPA firm whose agents run reconciliations, close, AP and AR, and…
- Aera Technology — Decision intelligence platform where an always on agent executes…
- Akro AI — On premise operational intelligence that automates document heavy…
- alfred_ — AI executive assistant that triages the inbox overnight and scores…
- Alloy.ai — Commerce intelligence system for consumer brands that unifies four…
Alternatives to Harvey
The closest documented capability profiles to Harvey among enterprise operations agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Glean12.0 / 14A lighter documented profile than Harvey
- Workato12.0 / 14A lighter documented profile than Harvey
- Boomi12.5 / 14Fuller documented coverage on Testing, Debugging & Optimization
- Coworker11.5 / 14A lighter documented profile than Harvey
- Fabrix.ai12.5 / 14Fuller documented coverage on Testing, Debugging & Optimization
- IBM watsonx Orchestrate12.5 / 14Fuller documented coverage on Testing, Debugging & Optimization
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded