OpenEvidence is the phenomenon of medical AI: valued at $12 billion (Series D, January 2026) and, according to the company, used daily by more than 40% of US doctors. But if you are a doctor or healthcare worker in Vietnam, the story is very different. This comparison is based on verified data, as of 11 July 2026.
What OpenEvidence is and where it shines
OpenEvidence is a point-of-care clinical question answering tool, free for US doctors verified with an NPI number, funded by advertising revenue aimed at healthcare professionals. Their real strength lies in exclusively licensed content: the full NEJM archive from 1990 (February 2025 agreement), JAMA plus 11 specialty journals (June 2025), Wiley (March 2026) and the NCCN cancer guidelines. This is something no competitor can replicate quickly, and it is the main reason investors value them so highly.
But there are three things users outside the US, including in Vietnam, need to know. First, full access requires an NPI, a US practice license number. Second, the company ended access in the EU and UK on 30 April 2026 over regulatory concerns, showing a complete focus on the US market. Third, on accuracy: the company reports a perfect USMLE score, but an independent study (medRxiv, December 2025) measured 34 to 41% on complex specialty questions, and a 2026 systematic review concluded the tool is strong on guideline-based questions but variable on complex cases. One consistent plus: it does not fabricate citations.
Quick comparison table
| Criterion | OpenEvidence | Phở Chat |
|---|---|---|
| Who can use it | US doctors verified with an NPI; exited EU/UK | Anyone; no practice-license gate, free plan available |
| Language | English, optimised for US practice | Vietnamese-native + English queries; correct medical terminology |
| Content sources | NEJM, JAMA, Wiley, NCCN (exclusively licensed) | PubMed, OpenAlex, arXiv, Europe PMC (open) |
| Answer style | Synthesis from licensed sources | Ranked by evidence tier, per-sentence cite-or-abstain ✅ |
| Drug interactions | Yes | Yes |
| Local guidelines | US guidelines | Vietnamese guideline integration on the roadmap 🟡 |
| Business model | Free, pharma advertising revenue | Freemium, paid plans, no ads |
| Bring your own API key (BYOK) | No | Yes: OpenAI, Anthropic, Google, DeepSeek, xAI keys, zero markup |
| Price | $0 (US doctors) | Free $0 · Starter $49.99/year · Pro $99.99/year |
✅ = shipped and live today · 🟡 = on the roadmap, not usable yet. We do not list unshipped features as if they exist.
When OpenEvidence is the right choice
You are a physician practising in the US with an NPI. Your questions follow US guidelines and you want NEJM and JAMA sources verbatim. In exactly that context, it is hard to find a better tool at zero cost.
Why Phở Chat fits doctors outside the US
The markets OpenEvidence chose not to serve are exactly where Phở lives. There is no NPI gate and no geo-exit: any clinician can sign up on the free plan. Answers rank sources by evidence tier, verify claims sentence by sentence, and stay transparent about anything that could not be verified. We do not run a pharma advertising model: the person paying is the user, so the product’s incentives point in only one direction, the correct answer. Cost stays under your control too: flat annual plans, or bring your own OpenAI, Anthropic, Google, DeepSeek or xAI API key (BYOK) with zero markup. Vietnamese clinicians get all of it natively in Vietnamese. Every clinical page in Phở carries the reminder: the tool assists with evidence lookup, treatment decisions belong to the physician.
An honest note: OpenEvidence figures are compiled from official press releases (PRNewswire, Businesswire), reputable press (TechCrunch, Fierce Healthcare, NBC News) and independent research (medRxiv, Research Square) as of 11 July 2026. Figures such as “40% of US doctors” are self-reported by the vendor. We compare to help Vietnamese readers choose the right tool, not to diminish a product that is excellent in its own market.