AI Agent Index
ByHeather MacAvelia·Independently reviewed·Published Apr 7, 2026·Updated Nov 8, 2026
Independently verified against live vendor data on Nov 8, 2026.

Enterprise legal AI platform whose Harvey Agents run legal work end to end. Raised $550M at a $15.5B valuation in September 2026. 200K+ professionals. SOC 2 Type II, FedRAMP Class D.

How we scored it

Autonomy

5/5

Integrations

5/5

Pricing clarity

2/5

Evidence

4/5

Setup

2/5

The facts

Pricing

Custom

GitHub

—

Stars

G2

4.2 / 5

35 reviews ↗

MCP

Server + client

Server + client

Reviews

Be the first →

Harvey is an enterprise legal AI platform founded in 2022 by Winston Weinberg, a former securities and antitrust litigator, and Gabriel Pereyra, a former research scientist at DeepMind and Meta AI. In September 2026 Harvey raised $550 million at a $15.5 billion valuation, co-led by Diffusion and Lightspeed Venture Partners, following a $200 million round at an $11 billion valuation in March 2026. Its backers include Sequoia, Kleiner Perkins, Andreessen Horowitz, GIC, Coatue, GV, and the OpenAI Startup Fund. Harvey states that more than 200,000 professionals at over 3,500 law firms and in-house legal teams in 85+ countries use the platform. Named enterprise clients include A&O Shearman, PwC, CMS, NBCUniversal, HSBC, Adecco Group, Dentsu, Cuatrecasas, Burges Salmon, Dentons, Reed Smith, KKR, Bridgewater, Deutsche Telekom, Procter and Gamble, Comcast, Merck KGaA, Vinson and Elkins, and Macfarlanes among others. The platform was named to the CNBC Disruptor 50 list in 2025 and again in 2026. Core products include Harvey Agents for end-to-end legal work execution, Assistant for document analysis and drafting, Vault for secure document storage and bulk analysis up to 100,000 documents, Knowledge for legal, regulatory, and tax research, Contract Intelligence for insights and negotiation support, Command Center for analytics and benchmarking, Harvey Mobile, Shared Spaces for cross-organization collaboration, Memory for carrying user preferences across Harvey, and the Ecosystem integration layer. Harvey II, introduced in August 2026, opens agents inside a matter or project Space with its documents, parties, permissions, ethical walls and history already in place, and Harvey Graph maps the relationships across a matter so agents share the same picture of the work.

Harvey integrates with the enterprise legal technology stack through the Ecosystem product. Document management integrations include iManage, NetDocuments, Google Drive, and SharePoint/OneDrive, letting users work with firm files inside Harvey and export results without downloading local copies. Research content from LexisNexis is available through a product partnership, alongside firm-specific precedent libraries, and an Intapp integration brings ethical-wall enforcement into Harvey. Word and Outlook add-ins bring Harvey into Microsoft Office, and Harvey is also available inside Microsoft 365 Copilot. The Connector Library, generally available since October 21, 2026 after an early-access period that began in June, holds connectors that each pass a Harvey security review, among them iManage, NetDocuments, Box, Everlaw, DocuSign, Datasite, Intralinks, PitchBook, Gmail, Slack, Microsoft Teams and Jira. Jus Mundi's arbitration research agent, Jus AI, works inside Harvey's Assistant, Agents, Vault and Workflows under a partnership announced on November 4, 2026. Harvey Agents operate across these connected systems to execute end-to-end workflows spanning document review, research, analysis, and drafting. The platform is designed around accessing Harvey where attorneys already work rather than requiring a switch to a new interface. Data can be hosted in the EU, Switzerland, the US, or Australia for data residency requirements. Harvey publishes a first-party MCP server documented at developers.harvey.ai, exposing five tools that cover general legal questions, knowledge-source research, Vault project listing and Vault document analysis, with setup guides for Claude Web, Claude Desktop, Google Gemini and Microsoft 365 Copilot over OAuth. Harvey also works as an MCP client: its Connector Library connects to partner MCP servers, firms can add custom MCP connectors for their own systems, and admins set governance policies for which tool actions are permitted. A public developer API is also published, with OpenAPI specifications for the Vault and completion endpoints.

Pricing is custom with no self-serve tier and no public pricing page, and firms must engage Harvey sales for a quote. Harvey does not publish a setup or implementation timeline, and no self-serve trial is offered on its site. It markets to mid-sized firms as well as large ones, but there is no plan a solo practitioner can buy online.

Harvey is designed for law firms and corporate legal departments buying on an enterprise contract. Teams focused mainly on contract review and drafting should also evaluate Spellbook. Plaintiff firms that want medical chronologies, demand letters, and mass tort screening should evaluate Eve. Firms already in the Thomson Reuters ecosystem may find CoCounsel ($360/mo billed annually for CoCounsel Essentials, with Westlaw included in the higher CoCounsel Legal plan) provides tighter integration with existing Westlaw research workflows.

Current state Q4 2026: Harvey announced its $550 million round at a $15.5 billion valuation on September 9, 2026, alongside its first post-trained open-weight model, the launch of Harvey LAB, its Legal Agent Benchmark, and Guardrails AI joining Harvey. On October 8, 2026 it announced TX-RAMP Level 2 certification, effective September 22, 2026, which opens the platform to Texas state agencies, state courts and public universities. On October 28, 2026 it announced FedRAMP Class D certification: Harvey runs within Knox Systems' FedRAMP High managed platform, where the FedRAMP Marketplace lists it as Harvey for Government. Its trust center lists SOC 2 Type 2, ISO 27001, ISO 27701, ISO 42001, AIUC-1, FedRAMP Class D, IRAP, TX-RAMP, GDPR and CCPA, and Harvey contractually guarantees through its Platform Agreement that customer data is not used for model training. External audits and testing are performed by firms including Schellman, NCC Group, and Bishop Fox. Harvey acquired Aida on October 29, 2026, its fifth acquisition of the year after Benchmark in July and Guardrails AI in September, and on October 30 it added Italian legislation, case law and commentary from the publishers Simone, Il Foro Italiano and Juranet. At Harvey FORUM Chicago on November 5, 2026 it put agents in the Outlook add-in, which triage email, draft replies and start first-pass review with the lawyer approving each step, and opened early access to Intake, which routes contract requests that arrive by email, and to a fact graph for litigation and investigations. A two-way iManage integration that sends finished work back to iManage is announced for early access, and Japanese legal research from Legalscape is still due in the coming months. CVS Health, Cozen O'Connor and Commercial Bank of Dubai announced in October and November 2026 that they had selected or deployed Harvey.

Pricing

custom

Segment

Enterprise

Setup

complex

Verified

Nov 8, 2026

Transparency

Quote Only

Contract

Not Published

Data training

Not Trained

Human in loop

Optional

Harvey AI pricing

Harvey AI uses custom pricing: the vendor quotes a price rather than publishing one. The vendor does not publish its contract terms. Checked against the vendor's own pages on Nov 8, 2026.

Harvey AI features

Deep researchData analysisCitationsAutonomous

Harvey AI pros and limitations

Editorial assessment

Pros

  • ✓Harvey Agents execute complex legal work end to end: from contract analysis and due diligence through drafting and research, agents plan the work, split large tasks across parallel agents, and return cited, review-ready output without step-by-step attorney direction. Scheduled Agents can also run recurring work in the background.
  • ✓Deep enterprise adoption and backing: Harvey states that more than 200,000 professionals at over 3,500 law firms and in-house legal teams use it, named clients include Dentons, KKR, Bridgewater, PwC, Deutsche Telekom, and Procter and Gamble, and it raised $550M at a $15.5B valuation in September 2026.
  • ✓Works where legal teams already work: iManage, NetDocuments, SharePoint, Google Drive and Box connections, Word and Outlook add-ins, LexisNexis research content, and MCP in both directions, with a first-party MCP server for outside assistants and MCP connectors that pull in PitchBook, Datasite and SS&C Intralinks data.

Limitations

  • –Enterprise-only pricing with no public price list and no self-serve plan, so every deployment starts with a sales conversation and buyers cannot compare cost before a demo. Spellbook and CoCounsel ($360/mo billed annually) are two alternatives that teams comparing legal AI platforms can evaluate.
  • –Harvey publishes no setup or implementation timeline and offers no self-serve trial, so firms cannot switch it on and evaluate it on their own before engaging Harvey sales. Several new capabilities are still in early access, including Intake for in-house contract requests and the litigation fact graph, and firm admins carry the governance work of choosing which connectors and tool actions are permitted.
  • –Independent review volume is small relative to adoption: the G2 and Gartner Peer Insights corpora each hold well under fifty reviews, against more than 200,000 professionals using the platform.

Harvey AI technical details

Deployment
webapimobileadd-in
Model architectureMulti-model: models from OpenAI, Anthropic, Google and Mistral (AI subprocessors on its trust center) run through Harvey, plus its own post-trained open-weight model announced in September 2026
Avg setup timeNot published. Harvey states no setup or implementation timeline, and deployment is sales-led with no self-serve trial.
Autonomous rateHarvey Agents plan and execute legal work end to end, split large tasks across parallel agents, and return cited, review-ready output, while Scheduled Agents run recurring work in the background. Attorneys review the finished output rather than approving each step. Harvey publishes no production autonomous completion rate. Its Legal Agent Benchmark reports rubric pass rates on held-out research tasks, which measure model performance rather than resolution of customer work.
Integrations
SlackGmailMicrosoft TeamsJiraGoogle DriveMicrosoft OutlookiManageEverlawDocuSignDatasiteNetDocumentsBoxSharePointOneDriveMicrosoft WordEponaIntappMimecastLawVuDocument DrafterDealCloserCourtroom InsightDocket NavigatorJudy LegalCourtAlertStrettoAnsaradaSS&C IntralinksLegaticsIdealsDeepJudgeMidpageRed Maple ResearchLexgraphThe Chancery Lane ProjectOctusPitchBookScopeLuplAderantAutologyxSyseroAffineSigParserMaryEntropiaNewcode.AIElevenFloAlizeStilta IP ABYoutrustGetCounselPointOneLaurelLexisNexisJus Mundi
Security
SOC 2 Type IIISO 27001ISO 27701ISO 42001AIUC-1IRAPTX-RAMPGDPRCCPAFedRAMP

Alternatives to Harvey AI

See all Harvey AI alternatives →

Rating

4.4/ 5
★★★★★

Editorial score

How we score this →

Recognition

MCP Server VerifiedListed 2026
Embed these on your site →

Industries

LegalPublic SectorHealthcarePharmaEnergyMediaTelecom

Built for

EnterpriseB2B

Leave a review

Used Harvey AI? Tell other buyers how it worked for you. Takes about two minutes.

Write a review

Compare

Related Content

Is this your product?

AI systems read this listing 108 times in the last 30 days. Last checked Nov 8, 2026. AI assistants repeat what this page says, so keep it current: Featured Listings are re-checked every 14 days.

Get Featured →Claim it free

AI Agent Price & Rating Tracker

Free, every two weeks. Price changes, new agent launches, acquisitions, and rating updates across the AI agent market. Verified against live vendor data, not vendor marketing.

No spam. Unsubscribe anytime. We never share your email. Read past issues