A $15 Billion Diagnosis: OpenEvidence's Quiet New Round Puts It on the M&A Exam Table

A company that most consumers have never heard of just became one of the most valuable vertical AI companies on Earth — and the people who built it are openly wondering whether to sell it.

OpenEvidence, the AI-powered clinical search engine often described as “ChatGPT for doctors,” has quietly closed a $250 million round at a $15 billion valuation, according to a Business Insider exclusive published September 24. The round, which includes investment from hospital systems and Andreessen Horowitz (with Sacra reporting Byers Capital as co-lead), marks a 25% step up from the $12 billion valuation set by its January 2026 Series D — itself a doubling from the $6 billion mark set just three months earlier. Two doublings in roughly nine months is a trajectory that even this frothy market rarely produces.

What actually happened

According to people familiar with the matter, the new round was closed recently and only surfaced this week through BI’s reporting. The numbers, cross-referenced with Sacra’s private-markets data, sketch an extraordinary arc:

• July 2025: $210 million round at a $3.5 billion valuation
• October 2025: $6 billion valuation, reportedly in a round that included Nvidia as an investor
• January 2026: Series D at $12 billion, led by Thrive Capital and DST
• September 2026: $250 million at $15 billion, with hospital systems and a16z participating

That means OpenEvidence has raised over $1 billion in the past year alone, from investors including Thrive, DST, GV, Kleiner Perkins, Sequoia Capital, and Nvidia. Founded in 2022 by Daniel Nadler — who previously sold Kensho to S&P Global for $550 million — the Cambridge, Massachusetts-headquartered company (with operations in Miami) has compounded valuations roughly 4.3x since the summer of 2025.

The commercial engine behind those numbers: Sacra estimates OpenEvidence hit approximately $300 million in annualized revenue by mid-2026, doubling in seven months. CNBC has reported that more than two-thirds of U.S. doctors rely on the product for diagnostic and treatment guidance — roughly 40% of all U.S. physicians by other estimates, making it plausibly the fastest-adopted professional software product in the healthcare industry’s history.

The twist: an exit that isn’t an IPO

What elevates this from a routine mega-round story is the second half of the BI report. Sources say OpenEvidence’s founders are fielding acquisition interest and could be open to selling the company rather than pushing toward a public listing.

One rationale is peculiar to this moment: the company’s product depends on access to computing power, and its leadership is reportedly weighing whether rising anti-AI sentiment could curb data center expansion in the coming years, making compute scarcer and more expensive. Buying a permanent seat at the compute table — by joining a hyperscaler or AI lab — is one way to hedge that risk. Business Insider could not verify specific M&A interest from named acquirers, and the company may well prefer to remain independent.

The strategic logic for an acquirer is obvious. OpenAI, Anthropic, and Google are all making a bigger push into healthcare, and increasingly they’re building healthcare tools of their own rather than just supplying models to other companies’ products. OpenEvidence offers a shortcut: doctors already using the product daily, hardened infrastructure for handling sensitive medical information, and distribution that would take a frontier lab years to replicate. Anthropic announced a partnership with the company just this week, and OpenEvidence already integrates with Google’s and Microsoft’s AI tools.

The M&A context matters too. The report lands amid a wave of consolidation: Nvidia has agreed to buy Hugging Face for nearly $13 billion, and Stripe has agreed to acquire OpenRouter for around $8 billion. If OpenEvidence does go up for sale, it would instantly become one of the largest AI acquisitions ever contemplated — a $15 billion entry ticket, plus a premium.

Why the numbers are defensible

Skeptics will note that a $15 billion valuation on $300 million of annualized revenue implies a 50x multiple. But vertical AI economics look different from horizontal SaaS. OpenEvidence’s product sits directly in clinical workflow, is used habitually by a majority of U.S. physicians, and monetizes through a mix of advertising, enterprise contracts with hospital systems, and — as of this round — direct investment from those same hospital systems, which now have skin in the game. Its moat isn’t the underlying model, which it doesn’t own, but the workflow lock-in, the trust of the medical profession, and the data relationships it has built with premier institutions.

The risks are equally real. The company is structurally dependent on frontier models it doesn’t control at a moment when OpenAI and Anthropic are building competing healthcare products — the equivalent of renting your engine from companies now building their own cars. Clinical AI carries permanent liability exposure. And a 4.3x valuation run in 14 months leaves little room for a growth stumble, particularly if the IPO window stays shut.

What to watch

Three signals will tell us where this ends. First, whether the reported sale process matures into named-acquirer talks — the usual suspects (OpenAI, Google, Microsoft, Nvidia, and the EHR incumbents) all have strategic reasons to want this asset. Second, whether the hospital systems that just invested gain governance rights or board seats, which would signal a co-opetition strategy rather than a pre-sale tidy-up. Third, whether OpenEvidence’s revenue keeps its seven-month doubling cadence into 2027 — the only metric that makes either the IPO or the M&A math work at these levels.

Either way, a company founded four years ago to help doctors search medical literature now sits at the intersection of every major force in AI: the compute squeeze, the healthcare land grab by frontier labs, and the great consolidation now reshaping the industry’s map. Whether it sells or not, the $15 billion diagnosis is in — and the patient is very much for examination.