Harvey's $550M round prices legal AI at 39 times the ARR it just crossed
Lightspeed and Diffusion co-led $550m at a $15.6bn mark, which added $4.6bn of paper value in six months and only reads as ordinary software if $400m of ARR becomes about $1.56bn. Eighty of the Am Law 100 already buy.
The Investor · Invest desk
What happened
- Harvey closed a $550 million round on September 9 at a $15.6 billion valuation, 41% above the $11 billion mark it printed six months earlier.
- Annual recurring revenue has crossed $400 million across more than 3,000 organizations, among them 80% of the Am Law 100, 20% of the Fortune 500 and five of the Fortune 10.
- Lightspeed Venture Partners co-led with Diffusion, a new firm co-founded by Kris Fredrickson, who had backed Harvey previously through Coatue Management.
- The valuation ladder runs $3 billion in February 2025, $5 billion four months later, $8 billion in December 2025 and $11 billion in March 2026 before this close.
Why it matters
- What it limits constraint Buying at roughly 39 times revenue commits the new investors to a near-quadrupling of ARR before the entry price resembles ordinary enterprise software, so the round is underwritten on growth rather than on holding the base.
- Who has to decide decision Spending the money on Harvey's own training and serving infrastructure rather than frontier API capacity puts a Beijing lab's open-weight release schedule inside Harvey's product roadmap.
- Who's exposed exposure The privilege case for owning the stack holds only while no frontier lab sells a comparable self-hosted tier to regulated buyers; once one does, self-hosting becomes a procurement requirement any vendor can meet.
- What it sets up precedent With 80 of the Am Law 100 already contracted, the next legal AI seller into that segment is making a displacement sale rather than an adoption sale.
Next story · Product · 8h ago
China's draft export list would cover the weights of its biggest AI models while leaving API access openConfirmed8 publishers
Divide $400m of annual recurring revenue by the more than 3,000 organizations now using Harvey and the average customer is paying somewhere near $133,000 a year [1][17], a real line item in a law firm's technology budget and a long way from the roughly $5.2m of valuation each of those logos is carrying [18]. Eighty of the Am Law 100 are in the base already [1][14], so the twenty that are not are the thinnest growth pool the company has.
The multiple is where the cash question sits. Techtimes puts the price at about 39 times ARR [9]; at a conventional ten times, the same $15.6bn mark corresponds to roughly $1.56bn of ARR, which is 3.9 times where Harvey is now [20]. The six months since the $11bn round produced $4.6bn of new mark against $550m of new cash, about $8.40 of paper value per dollar in [23][15], and the more than $1.5bn Harvey has raised since 2022 is now 9.6% of what the company is said to be worth [7][19].
The round has a job beyond the mark. Harvey Tenet downloads open-weight parameters, trains them on Harvey's own infrastructure, and runs the result inside an architecture that never sends client content to an outside provider [4], starting from Kimi K3, the model Moonshot AI released in July 2026 [5]. K3 is a 2.8-trillion-parameter mixture-of-experts that activates 16 of its 896 experts per token, 1.8% of them [6][16], which is what keeps the inference bill from tracking the parameter count. So Harvey is buying GPUs and a training team rather than API capacity from OpenAI, Anthropic or Google, and techtimes argues the reason is doctrinal rather than economic: privilege can be waived when privileged material reaches an unauthorized third party, and a model API provider is a candidate for that role [11], with courts treating self-hosted deployment as the strongest protection available [24]. Harvey's promise not to train on customer data [12] narrows the exposure without removing the third party from the chain.
What the evidence does not establish is that 39 times is now the reference price for vertical AI generally. The source offers no comparable multiple from any other vertical AI round, only the assertion that few enterprise software companies have compounded value this fast, and a ladder from $3bn in February 2025 through $5bn, $8bn and $11bn to here, a 5.2-fold move [10][8][21]. That is a claim about scarcity of comparables, not evidence of a benchmark.
The risk I would underwrite against is not the moat failing but the moat holding while growth changes shape: with 80% of the Am Law 100 signed and 20% of the Fortune 500 [1], the next tranche of revenue comes from in-house legal departments and from selling more seats and workloads into firms Harvey already serves, and seat expansion inside partnerships whose compensation still runs on billable hours is slower arithmetic than new-logo land. Winston Weinberg, a former O'Melveny litigator, presumably knows that better than his investors do [13]. Priced here, Harvey has taken on the obligation to add about $1.16bn of ARR before its own entry price reads as ordinary software [22].
What to watch
- Whether the remaining 20 Am Law 100 firms sign with Harvey or buy a competing self-hosted product.
- Any court ruling that treats a self-hosted legal model differently from a third-party enterprise API contract on privilege.
- Whether Tenet moves from the August 20 research preview into production work at named firms rather than staying a demonstration.
Information, not investment advice. This story is general information about companies and markets, drawn from cited reporting. It is not a recommendation to buy, sell or hold any security; check the cited sources before acting.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence55
- Adoption70
- Hype gap+30
- Incentives60
- Confidence60
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Harvey's annual recurring revenue has crossed $400 million and its customer base has grown to more than 3,000 organizations since March, including 80% of the Am Law 100, 20% of the Fortune 500, and five of the Fortune 10.
- [2]
The round was co-led by Lightspeed Venture Partners and Diffusion, a new firm co-founded by Kris Fredrickson, a longtime Harvey backer who previously invested through Coatue Management.
- [3]
Returning investors including Sequoia Capital, Andreessen Horowitz, Kleiner Perkins, Coatue, Goldman Sachs Alternatives, GIC, Verified Capital and WndrCo participated alongside new backers Sapphire Ventures and Whale Rock Capital Management.
- [4]
Harvey Tenet downloads a set of open-weight model parameters, trains them on Harvey's own infrastructure, and runs the resulting model inside an architecture that never transmits client content to an outside provider, rather than calling an API that routes documents through OpenAI, Anthropic or Google servers.
- [5]
The Tenet model, previewed in an August 20 research post, begins from Kimi K3, the open-weight foundation model released in July 2026 by Moonshot AI, the Beijing-based lab.
- [6]
Kimi K3 is a 2.8-trillion-parameter Mixture-of-Experts architecture that activates only 16 of 896 expert sub-networks per token, keeping inference costs far below what the full parameter count would suggest, and supports a one-million-token context window.
- [7]
Harvey has raised more than $1.5 billion since its founding in 2022.
- [8]
Harvey was worth $3 billion in February 2025, $5 billion four months later after a Kleiner Perkins-led $300 million round, $8 billion by December 2025 after an Andreessen Horowitz-led round, and $11 billion in March 2026 after $200 million from Sequoia and GIC.
- [9]
Techtimes states that at $400 million in ARR the $15.6 billion valuation implies a revenue multiple of roughly 39 times, elevated by conventional enterprise software standards but in line with what investors will pay for a company they believe is defining a new software category.
- [10]
Techtimes states that few enterprise software startups have compounded their valuation as quickly as Harvey, describing a more than fivefold increase in roughly 18 months.
- [11]
Attorney-client privilege can be waived when privileged material is transmitted to an unauthorized third party, and when a firm routes client documents through a third-party AI model API the API provider becomes a potential unauthorized recipient.
- [12]
Harvey announced that it will not use customer data in training.
- [13]
Harvey was founded in 2022 by Winston Weinberg, a former securities and antitrust litigator at O'Melveny & Myers, and Gabe Pereyra, previously a researcher at Google DeepMind and Meta, after they experimented with early versions of GPT-3.
- [14]
80% of the Am Law 100 is 80 firms, leaving 20 of that group not yet customers.
- [15]
The move from $11 billion to $15.6 billion added $4.6 billion of valuation over six months against $550 million of new capital, about $8.40 of new mark per dollar raised.
- [16]
Activating 16 of 896 experts per token means about 1.8% of Kimi K3's experts fire on any given token.
- [17]
At more than $400 million of ARR across more than 3,000 organizations, average revenue per customer organization is about $133,000 a year.
- [18]
The $15.6 billion valuation spread over more than 3,000 customer organizations is about $5.2 million of valuation per organization.
- [19]
Cumulative capital raised of more than $1.5 billion equals about 9.6% of the $15.6 billion valuation.
- [20]
For the $15.6 billion mark to correspond to a conventional ten times revenue multiple, ARR would need to be about $1.56 billion, which is 3.9 times the current $400 million.
- [21]
From $3 billion in February 2025 to $15.6 billion, the valuation is up 5.2-fold.
- [22]
Reaching $1.56 billion of ARR from $400 million requires adding about $1.16 billion of ARR.
- [23]
Harvey, the San Francisco-based legal AI company, closed a $550 million funding round on September 9 that values it at $15.6 billion, a 41% jump from the $11 billion mark set six months earlier.
- [24]
Courts have recognized that privilege protection varies depending on whether AI is deployed via a consumer tool, a third-party enterprise contract or self-hosted infrastructure, with self-hosted deployments offering the strongest protection.
Sources
2 independent publishers whose own reporting we read for this story.
- pivotnews.aiHarvey raises $550M at a $15.5B valuation, SiliconANGLE reports
1 article · September 10, 2026
- techtimes.comHarvey Closes $550M for Legal AI Model That Keeps Client Documents in Firm Control
1 article · September 9, 2026
Topics and entities
Topics
- Attorney-client privilege and AI
- Open-Weight Model Publishing
- AI Startup Funding
- Legal AI