

OpenEvidence is a medical question-and-answer tool for clinicians. You ask a clinical question and get a cited answer drawn from peer-reviewed sources. It is free and unlimited for verified US healthcare professionals, and it is paid for by ads.
The company calls itself the most widely used medical AI among verified US clinicians. It claims more than 200 million consultations to date. This review covers what it does, who can get in, and where the catch is.
The core is the ask bar. Type a clinical question, in English or another language, and OpenEvidence answers with citations you can open. The sources are licensed, not scraped. OpenEvidence has content agreements with the New England Journal of Medicine, JAMA, Nature, Cochrane and the NCCN cancer guidelines. It names more than a dozen medical societies as partners.
Each answer carries an EvidenceGrade. That is a label showing how strong the cited evidence is. A single case report does not read like a trial. You can also ask it to build a workup for a presentation.
Around the ask bar sit clinical tools. Visits records a patient encounter, transcribes it, and drafts the note with evidence pulled into the assessment and plan. Coding Intelligence suggests CPT and ICD-10 codes and an E/M level with the reasoning written into the note. Doctor Dialer sends texts, faxes and voicemail from a dedicated number. Patient Take-Homes turn the visit into plain-language material for the patient.
There are iPhone and Android apps. Since September 2026 it runs on OpenEvidence's own models, led by one named Darwin.
Nothing. There is no paid plan, no usage cap and no premium tier. The company says it is committed to keeping the service free and equal for every healthcare professional.
4 Views
The money comes from advertising. OpenEvidence shows what it calls a limited number of controlled ads from vetted advertisers. Its policy states the ad system and the answer system are separate, and that advertisers cannot influence answers. The ads are marked as ads.
Verified US healthcare professionals. Sign-up checks your credentials before the platform opens. Doctors, nurse practitioners, physician assistants and other licensed clinicians qualify. Medical students and trainees have access too.
If you are not a US clinician, you are locked out. There is no patient version and no general-public tier.
It has one now. Visits, launched in August 2025, records the encounter and drafts the note. It adds evidence and guidelines into the plan as you go. You can also query all of a patient's uploaded documents.
It is a scribe attached to a search engine, though, not a scribe first. Dedicated scribes such as Freed or Heidi offer EHR push and specialty templates as the main event. OpenEvidence does not publish an EHR integration list. If notes are the whole job, compare before you switch.
US only. Verification is built around US credentials, and every partner society and guideline is American. Clinicians elsewhere cannot register.
Ad-funded medicine deserves a raised eyebrow. The separation policy is clear and public, but the advertisers are in the same industry as the answers. Judge that for yourself.
The benchmark claims are the company's own. A perfect MedQA score and top marks on HealthBench come from OpenEvidence's announcement, not from an outside audit.
There is no offline mode and no API for building on top of it. The product is the app.
The homepage blocks browsers without JavaScript and does not publish a plain FAQ or pricing page. Basic questions are answered only in announcements and the user guide.
Yes, if you can get in. For a US clinician the price is zero and the sources are the ones you already trust. The ask bar alone earns its place on a phone.
Treat Visits and the coding tools as a bonus rather than the reason to sign up. And keep the ad question in mind. The answers are cited, so check the citations, which is what the tool is for.