OpenEvidence: Medical AI Search and Its Sources

OpenEvidence is an AI search tool for doctors and other health care professionals. It answers clinical questions from licensed medical literature and links to the sources it used. OpenEvidence does not document a web crawler, so robots.txt has no documented effect on it. Site owners have nothing to allow or block.

Who uses it

The OpenEvidence about page describes the product as an AI copilot for doctors at the point of care. The team is led by Daniel Nadler. Users log in and are verified as health care professionals, as the NEJM announcement describes. The tool is available in web browsers and in apps for iOS and Android.

Where answers come from

OpenEvidence lists content agreements with NEJM, the JAMA Network, Nature, NCCN and Cochrane on its about page.

The NEJM agreement is an example of what these deals cover. On February 19, 2025, OpenEvidence and NEJM Group announced a multi-year agreement. All content and multimedia from 1990 forward from NEJM, NEJM Evidence, NEJM AI, NEJM Catalyst and NEJM Journal Watch inform OpenEvidence answers. Answers also use recent invited reviews and selected figures and tables from NEJM.

A resource review in the Journal of the Medical Library Association describes the method. OpenEvidence uses a retrieval-augmented generation (RAG) model. RAG means that the model writes its answer from documents it retrieved for the question. Each answer has citation links, so the user can open the source.

Web crawling and opt-out

OpenEvidence does not publish a crawler name, a user agent string or a robots.txt token. Its documented sources are licensed journals and clinical guidelines, not the open web. A general website cannot opt in or opt out of OpenEvidence through robots.txt.

OpenEvidence names its agreements with publishers as the source of its content. It does not describe a way for other sites to submit content.