MIAMI, January 22, 2026, 13:55 (EST)
- OpenEvidence secured $250 million in funding, boosting its valuation to $12 billion—twice what it was just three months ago
- The company said Thrive Capital and DST Global co-led the Series D round.
- The startup claims over 40% of U.S. physicians use its tool every day
OpenEvidence, a medical AI startup with tools in use by U.S. doctors, announced it has secured $250 million in new funding, pushing its valuation to $12 billion—twice what it was just three months ago. The Series D round, co-led by Thrive Capital and DST Global, brings the company’s total funding close to $700 million. Back in October, OpenEvidence was valued at $6 billion after raising about $200 million, according to PitchBook. The company claims its platform is now used daily by over 40% of U.S. physicians across more than 10,000 hospitals and medical centers. (Reuters)
The round’s size is significant since healthcare has been a more cautious, slower adopter of generative AI—systems that generate human-like text—compared to consumer apps or office tools. Clinicians need to trust the information they get; errors could end up in a patient’s chart, prescription, or referral.
This bet hinges on the idea that “specialized” outperforms “general” in healthcare. While investors flood broad chatbots trained on the open web with cash, hospitals prefer tools that can back up their answers, cite sources, and rely strictly on vetted medical content.
OpenEvidence markets itself as a clinical evidence search tool rather than a typical chatbot, relying on peer‑reviewed journals and clinical guidelines to provide answers and relevant citations. The company highlights partnerships with organizations like the New England Journal of Medicine and the American Medical Association to bolster its claims of accuracy and reliability.
The company reported backing roughly 18 million clinical consultations from verified U.S. physicians last December, a jump from about 3 million monthly consultations a year before. It intends to use the fresh funding to boost research and development and expand an AI system that directs doctor inquiries to more specialized medical platforms.
OpenEvidence describes its product as a “brain extender” that provides citation-linked answers drawn from peer-reviewed studies. The company plans to use the new funding for compute power and a “multi-agent” architecture—a system that routes queries to various specialized models before combining the results. CEO Daniel Nadler noted that staying current with new evidence and guideline updates could consume “nine hours of their day, each day.” Meanwhile, Thrive partner Kareem Zaki dubbed the product “the default operating system of medical knowledge” in the U.S. (Business Wire)
STAT News reported that the company is available at no cost to clinicians who have a National Provider Identifier, the U.S. ID for healthcare providers, and generates most of its revenue through ads targeted at clinicians. The outlet noted OpenEvidence secured $735 million in funding over the last year, while Nadler has claimed the $12 billion valuation signals potential to grow beyond advertising. (STAT)
OpenEvidence is among a rising number of medical AI companies aiming to embed themselves into everyday clinical tasks — like search, note-taking, and triage — without bogging down appointments. Their pitch boils down to this: spend less time digging for evidence, and more time with patients.
But the risks are clear. Even AI trained on reliable data can confidently produce mistakes. Hospitals remain cautious, concerned about liability, data privacy, and whether the model’s responses are consistent across specialties and diverse patients. If clinicians don’t trust the results, or if health systems see the risks outweighing the benefits, adoption could grind to a halt.