Skip to content

Funding Brief|Harvey $550M, $15.5B Valuation

Sep 10, 20261 min
TL;DRHarvey closed a $550M round co-led by new investors Diffusion and Lightspeed, pushing its valuation from $11B in March to $15.5B — nearly doubling in nine months. Half the money buys confidence for its own fine-tuned model, half buys agent-security capability via the Guardrails AI acquisition: Harvey is shifting from a law-firm AI assistant into a vertical AI company that trains its own models and manages its own agent risk.
Table of Contents
  1. Funding Details
  2. What the company does
  3. What this round signals
    1. What it means for the agent ecosystem
    2. What investors are betting on
    3. Numbers worth watching
  4. Watchlist status
  5. Today's takeaway
  6. References

🌏 中文版

Funding Details

FieldValue
CompanyHarvey (San Francisco, USA)
RoundUnnamed round (TechCrunch's read: at least 8 priced rounds to date, two of them extensions, so this one could be considered a Series F)
Amount$550M
Lead investorsDiffusion, Lightspeed Venture Partners (both new investors)
Follow-onSequoia, Kleiner Perkins, Andreessen Horowitz, Coatue, Conviction, Elad Gil, Evantic, GIC, Goldman Sachs Alternatives, Verified Capital, WNDR (existing investors), plus new investors Sapphire Ventures and Whale Rock Capital Management
Valuation$15.5B (up from $11B in March, roughly 41% in 9 months; was $8B as of December 2025)
Total raisedOver $1.55B
Founded2022
HeadcountRoughly 860-1,260 (sources diverge: Revelio Labs puts Q1 2026 headcount at 1,261; Silicon Valley Investclub lists 860+)

What the company does

Harvey is a vertical AI platform for the legal industry — handling contract analysis, due diligence, regulatory compliance, and litigation research for law firms and corporate legal departments. Its customers now include 80% of Am Law 100 firms and legal teams at five Fortune 10 companies.

Its core product line spans contract and document analysis tools plus Vault, an enterprise knowledge-base integration layer. In August it shipped its first fine-tuned model, Harvey Tenet, built by post-training Moonshot AI's open-weight Kimi K3 with help from inference provider Fireworks — the same provider that helped Cursor train its own homegrown model. Harvey is also encouraging customers to adopt and post-train their own open-weight models, a deliberate move away from locking into a single frontier-model vendor — a relatively unusual stance in a legal AI space that leans heavily on OpenAI and Anthropic.

Customer count grew from 1,300 in March to over 3,000 today, and annual recurring revenue has crossed $400M. The round also closed the acquisition of Guardrails AI, a platform for testing AI agent security and behavior — Harvey's fourth acquisition this year.

What this round signals

What it means for the agent ecosystem

The Guardrails AI acquisition is more telling than the dollar amount. Harvey is dispatching more agents to handle long-running, autonomous legal work, and Guardrails AI exists to test exactly that: how safely an AI agent behaves. This marks legal AI reaching a stage where agents make extended autonomous judgment calls on clients' sensitive documents and litigation strategy — agent security testing has moved from optional to acquisition priority. The new capital will also fund senior engineering hires; CEO Winston Weinberg has signaled Harvey will keep filling capability gaps through acquisitions rather than building everything in-house.

What investors are betting on

Diffusion and Lightspeed came in leading straight away, while existing backers — Sequoia, a16z, Coatue among them — piled in alongside them. That "new and old money doubling down together" pattern signals no disagreement in the market about Harvey's growth trajectory. Sequoia has backed Harvey since 2023, and this follow-on fits its usual vertical-AI playbook (legal, coding) of betting early on people and adding capital as growth compounds. Notably, this round's valuation jump (41%) is slower than the pace before it — the March round went from $8B to $11B (37.5%) in just three months — suggesting legal AI valuations may be shifting from an "explosive" phase to a "steady climb."

Numbers worth watching

  • ARR crossed $400M, up from an undisclosed figure in March, while customer count grew from 1,300 to over 3,000 in the same window — growth above 130%
  • $15.5B valuation against $400M+ ARR implies roughly a 38x multiple — a touch below Cognition's ~53x in the same period, but above the typical 15-25x for late-stage SaaS
  • Three valuation jumps in nine months ($8B → $11B → $15.5B), a cumulative gain near 94%, showing capital heat in legal AI hasn't cooled

Watchlist status

Harvey is already tracked in watchlist section D4 (Legal AI). Tracking focus updated to: $550M new round, $15.5B valuation, simultaneous acquisition of AI agent security startup Guardrails AI, in-house model Harvey Tenet now live.

Today's takeaway

The more interesting part of this round isn't the number — it's that Harvey casually picked up an agent-security startup along the way. As a company's core product evolves from "AI-assisted lawyer" to "agent autonomously handling legal work," security testing stops being a nice-to-have and becomes infrastructure you can't ship without. That's the same signal as Cognition rolling out Devin Security Swarm alongside its Series E: two agent categories that grew up separately — coding agents and legal agents — are both folding agent security into their own product lines instead of waiting for an outside security startup to supply it.

References