ChatGPT vs WebMD Symptom Checker: the AI chat era vs the checklist era
Last updated September 3, 2026.
The way people check symptoms has split into two eras: the checklist era, owned by WebMD's Symptom Checker, and the chat era, where people type their symptoms into ChatGPT. OpenAI leaned into this with ChatGPT Health, a dedicated health experience it announced for conversations about medical topics, per its own announcement. So which tool actually serves you better when something hurts at midnight?
How do they work, and what do they cost?
Both are free at the entry level. WebMD's Symptom Checker is a structured checklist: pick symptoms on a body map, select multiple symptoms, get possible conditions and treatments, with a beta version in testing, per its own site. ChatGPT is a conversation: describe symptoms in your own words, get a discussion of possibilities, follow-up suggestions, and explanations of any term you do not understand, with ChatGPT Health framing the experience around medical conversations specifically.
What does the evidence say?
Be honest about both. A widely cited 2015 BMJ study of symptom checkers found they named the correct diagnosis first only about a third of the time, so the checklist era's accuracy was always modest. Chat models are more fluent and more flexible, but fluency is not accuracy: they can state wrong things confidently, and they know nothing about your history unless you type it in. Neither tool examines you, orders a test, or takes responsibility for being wrong, and WebMD says directly on its tool to call your doctor or 911 in an emergency.
So what should you actually use?
Use either for orientation, and neither for verdicts. WebMD's checklist is fast for simple, common symptoms. ChatGPT is better for explaining, translating jargon, and helping you frame questions for a real appointment. The most useful evolution of the idea is a purpose-built consult like Pymander's free AI doctor consult, which is designed around the clinical conversation and hands you a care note for a real doctor. Emergencies bypass all of it: call 911. Pymander routes emergency red flags to 911, and its escalation logic is documented in its safety architecture working paper.
What a Pymander AI doctor consult looks like
Illustrative example, not a real member's messages.
Common questions
Is ChatGPT reliable for health questions?
It is useful for orientation, explanation, and framing questions for a real doctor, but it can state wrong things confidently and knows nothing about you beyond what you type. OpenAI introduced ChatGPT Health as a dedicated experience for medical conversations, but no chatbot diagnoses you.
How accurate are symptom checkers like WebMD's?
Modest. A widely cited 2015 BMJ study found symptom checkers named the correct diagnosis first only about a third of the time. Treat the output as a possibilities list, never a verdict.
What is ChatGPT Health?
A dedicated health experience OpenAI announced for medical conversations, acknowledging that health questions are one of the most common things people ask ChatGPT, per its own announcement.
Is WebMD Symptom Checker free?
Yes. It offers a body map, multi-symptom selection, and possible conditions and treatments at no cost, with a beta version currently in testing, per its own site.
What should I use for a symptom that is worrying me?
Any of these tools for orientation, then a real clinician for anything persistent, worsening, or frightening. Pymander's free AI doctor consult is built for exactly this step: a clinical-style conversation that ends with a care note for your real care team.
What if it is an emergency?
Skip every app and call 911. WebMD says the same on its own tool, and Pymander is built to recognize emergency red flags and direct you to call rather than continue.