The $15bn ‘viral’ AI start-up changing how American doctors treat patients

Miami-based OpenEvidence, founded in 2022 by serial entrepreneur Daniel Nadler, says nearly two-thirds of US doctors use its AI search engine. The chatbot scours peer-reviewed journals (including NEJM and JAMA content) and clinical guidelines. Physicians, out of America's roughly 1mn, entered more than 42mn prompts last month. 'We're the first viral consumer internet app in healthcare,' says Nadler. It has raised nearly $1bn in the past 18 months from a16z, Blackstone, Thrive Capital and Sequoia, most recently $250mn in September at a $15bn valuation. Mayo Clinic is also an investor. It is free to doctors, funded by ads, and grew by word of mouth. According to a person close to the company, it had booked $100mn of ad buys (current plus future commitments) for the year by end-2025 and is on track for $400–500mn from ads by end-2026. Hospitals including Mount Sinai and Memorial Sloan Kettering are embedding it in electronic health records, and NYU Langone sought a formal tie-up because its clinicians were already using it. OpenEvidence is expanding into ambient documentation and clinical trials (cancer, rare diseases). It has hired Alexander Tarashansky, creator of Claude Science at Anthropic, and Nadler says the 'end game is literature-based drug discovery'. The eight senior physicians the FT interviewed praised it. Competitors include OpenAI's ChatGPT for healthcare, Anthropic's health tools (although Anthropic partnered with OpenEvidence for 100 low- and middle-income countries) and Wolters Kluwer's AI UpToDate (October 2025). Caveats: critics question whether an ad model justifies the valuation. OpenEvidence withdrew from the EU and UK earlier this year over the EU AI Act. A Mayo Clinic study found accurate, evidence-based answers in all five sampled cases but limited impact on treatment decisions, and a June Nature paper found it underperformed general models (Gemini, GPT) on clinical benchmarks. NYU Langone's Eric Oermann points to a lack of research on individual AI tools, and there is concern that junior doctors over-rely on it.