OpenEvidence
An AI clinical copilot that answers point-of-care medical questions from peer-reviewed literature, free to verified physicians.
1. Core Product / Service
OpenEvidence is a generative-AI medical search engine — a "copilot for doctors" — that answers clinical questions at the point of care and returns transparent, cited responses drawn from peer-reviewed medical literature. Its design thesis is that smaller models trained narrowly on vetted clinical sources outperform large general-purpose LLMs for medicine: the system uses an ensemble of specialized models, restricts training to peer-reviewed literature, and surfaces source citations for every answer [openevidence.com/about; Sequoia podcast, 2026-06-29].
The corpus is anchored by exclusive multi-year content licenses with the field's most authoritative publishers. A February 2025 agreement with NEJM Group provides all NEJM, NEJM Evidence, NEJM AI, NEJM Catalyst, and NEJM Journal Watch content from 1990 forward; a June 2025 deal with the JAMA Network adds JAMA and its eleven specialty journals [openevidence.com/announcements/openevidence-and-nejm; fiercehealthcare.com, 2026-06-29]. The company has also extended into hands-free voice query and EHR-embedded workflows (e.g., an Epic integration via Sutter Health) [fiercehealthcare.com, 2026-06-29].
2. Target Users & Pain Points
The user is the individual clinician, not the hospital IT department. OpenEvidence deliberately treats doctors as consumers — free signup for verified clinicians — bypassing the multi-year procurement cycles that gate most health-IT sales. The pain it solves is the impossible volume of medical literature: a physician cannot read every new trial, and general-purpose chatbots hallucinate and cannot be trusted at the bedside. OpenEvidence offers fast, cited, literature-grounded answers during a live patient encounter [Sequoia podcast; research.contrary.com, 2026-06-29].
Adoption figures are extreme for clinical software: ~760,000 registered U.S. physicians and ~18M clinical consultations/month as of December 2025, rising to ~20M/month and over 1M consultations in a single day by early 2026; the company states ~40% of U.S. physicians log in daily across 10,000+ hospitals [Wikipedia; fiercehealthcare.com, 2026-06-29].
3. Competitive Landscape
The point-of-care reference market has long been held by subscription incumbents built on human-authored summaries. OpenEvidence's wedge is a free, AI-native, advertising-funded model against $500+/year paywalls.
| Company | Model | Pricing | Notes |
|---|---|---|---|
| OpenEvidence | AI-native, citations from licensed journals | Free to clinicians (pharma-ad funded) | ~40% of U.S. physicians daily; NEJM + JAMA licenses |
| UpToDate (Wolters Kluwer) | Human-authored + new gen-AI layer | ~$500+/yr individual; institutional | Long-dominant incumbent; AI confined to its own dataset |
| DynaMed (EBSCO) | Human-curated evidence summaries | Subscription | Legacy point-of-care reference |
| ClinicalKey (Elsevier) | Reference / search over Elsevier corpus | Subscription | Publisher-owned distribution |
| ChatGPT / general LLMs | General-purpose, no clinical sourcing guarantee | Free / consumer paid | Used ad hoc by clinicians; reliability/hallucination concerns |
| abridge | Ambient scribe; added UpToDate/NEJM/JAMA CDS tie-ins | Enterprise per-seat | Adjacent — documentation, now expanding into decision support |
Differentiation: exclusive top-journal licenses, citation transparency, and a free distribution model that turns physicians into a direct consumer base rather than an enterprise sale.
4. Unique Observations
- The real business is not search — it is the highest-CPM advertising audience in existence. The ~600,000 U.S. prescribers command pharma-ad CPMs of $70–150+ versus $5–15 on consumer social, monetized via ads served during answer generation. This makes OpenEvidence structurally a pharma-marketing media company wearing a clinical-tool skin [medcitynews.com; Sacra, 2026-06-29].
- Distribution-first, not model-first. Where peers like hippocratic-ai and ambience-healthcare sell into health systems, OpenEvidence won the clinician directly and for free — arguably the fastest clinical-software adoption since Google. Owning the daily-active physician is the moat; the model is replaceable, the audience is not.
- The looming model shift: EHR embedding (Epic/Sutter) points toward enterprise per-seat pricing that could lift ARPU 5–10x over ads alone — a hedge against ad-only dependence and the obvious next leg of the $12B valuation thesis [iatrox/healthcare.digital, 2026-06-29].
- Daniel Nadler has done this shape before: Kensho applied AI to a high-value professional audience (finance) and sold to S&P Global. OpenEvidence is the same template aimed at medicine.
5. Financials / Funding
- Total raised (primary equity): $0.73B
- Latest valuation: $12.0B
| Date | Round | Amount | Post-money | Lead investor(s) |
|---|---|---|---|---|
| 2022 | Pre-A / early (undisclosed) | undiscl. | — | — |
| 2025-02 | Series A | $0.07B | $1.0B | Sequoia Capital |
| 2025-07 | Series B | $0.21B | $3.5B | GV (Google Ventures); Kleiner Perkins |
| 2025-10 | Series C | $0.20B | $6.0B | GV (Google Ventures) |
| 2026-01 | Series D | $0.25B | $12.0B | Thrive Capital; DST Global |
6. People & Relationships
- Founders / key people: Daniel Nadler (founder & CEO) — Harvard Ph.D., previously founder of financial-AI firm Kensho (acquired by S&P Global in 2018); named to the Time 100 Health list. Travis Zack serves as Chief Medical Officer [Wikipedia; Sequoia podcast; pearhealthcareplaybook, 2026-06-29].
- Notable investors: Sequoia Capital (Series A lead), GV / Google Ventures (Series B & C lead), Kleiner Perkins, Thrive Capital and DST Global (Series D leads) [Sacra; statnews.com, 2026-06-29].
- Content partners: NEJM Group (New England Journal of Medicine) and the JAMA Network — exclusive multi-year licensing deals.
- Distribution / health-system partners: Sutter Health and Cedars-Sinai (EHR integration and hospital deployment) [fiercehealthcare.com, 2026-06-29].
- Competitors: UpToDate (Wolters Kluwer), DynaMed (EBSCO), ClinicalKey (Elsevier), and general LLMs; adjacent to abridge and ambience-healthcare in the clinical-AI stack.