A free AI doctor for dads, a text away
Last updated October 19, 2026.
Pymander is a free AI doctor that works entirely by text message. No insurance, no appointment, no explaining to a receptionist why you finally called. You text, the consult begins.
Pymander is a free AI doctor that dads reach by text message: no insurance, no account, available 24/7 in the US for adults 18 and over. You describe what is wrong, it asks the questions a doctor would ask, and you get an assessment and a plan in minutes. It is a medical AI system, not a licensed doctor.
Key facts
| Price | Free, with or without insurance |
| How you reach it | Text or iMessage to +1 (689) 528-6964 |
| Insurance | Not taken. Nothing is billed, so nothing appears on an insurance statement |
| Availability | 24/7, including nights, weekends, and holidays |
| Who it is for | Adults 18 and over in the US |
| What it is | A medical AI system, not a licensed doctor |
| Published evidence | 98.2% on the USMLE across two independent runs; HealthBench Hard mean 0.3331 (N=999) on a sealed, pre-registered run, top five against every published result |
The short version: Pymander is a free AI doctor that works entirely by text message. A first consult starts the moment you text +1 (689) 528-6964. Pymander is a medical AI system, not a licensed doctor: it does not diagnose or treat, and it tells you plainly when you need a human clinician.
Why dads use it
No ceremony. A lot of dads will Google a symptom for a month before they will call a doctor once. Texting skips the phone call, the hold music, and the waiting room, and it still ends with a real assessment.
The "probably nothing" check. Most of the time the honest answer is "probably nothing, and here is what would make it something." Getting that answer takes minutes, and it replaces weeks of low-grade wondering.
It fits between things. At work, at the game, at 11pm after everyone is asleep. The consult is a text thread: it moves at your pace and picks up where you left off.
What you get and what you do not get
| You get | You do not get |
|---|---|
| A full consult by text: what this is likely to be, how worried to be, and what to do next | Prescriptions (they legally require a licensed clinician) |
| A red-flag check on every consult, with direct escalation when something needs emergency or same-day care | A formal diagnosis |
| Memory between visits: your history carries forward, so each consult starts where the last one ended | Labs or imaging orders (planned, with licensed clinicians reviewing) |
| A care note you can keep or share, viewable through a signed link | Emergency care itself. In an emergency, call 911 first |
How a consult works
- Text +1 (689) 528-6964. Say what is going on in your own words. There is nothing to install and nothing to sign up for.
- Answer a few questions. The AI doctor takes a real history the way a physician does, asking the one question that best discriminates between the possibilities it is weighing.
- Get your assessment and plan. What it most likely is, what to do tonight, what to watch for, and whether you need a human doctor, urgently or not. Your care note arrives as a link you can keep or share.
Built for the "I'll deal with it later" approach
Later is where symptoms go to get worse. The consult is deliberately the lowest-friction step that still counts as dealing with it: a careful history, the discriminating questions a doctor would ask, and a plan you can act on. If the answer is "see someone," you get that in plain words, with a care note to bring along so the visit starts ahead.
The published evidence
Pymander shows its homework. The clinical AI scored 98.2% on the USMLE, the exam human doctors take, across two independent, fully logged runs of the official sample examinations. On HealthBench Hard, an open-ended clinical-conversation benchmark, Pymander ran a sealed, pre-registered evaluation of 1,000 conversations and scored a mean of 0.3331 (N=999), top five against every published HealthBench Hard result. The protocol, generations, and judge grades are public and rerunnable at zenodo.org/records/22714759.
Safety routing is deterministic: a separate pipeline, not the model alone, decides when a situation escalates, and the architecture is published in full in the safety architecture working paper on the research page. It was evaluated on a 300-case red-flag suite, scoring 1.0 on the 60-case gate sample with zero under-escalation on the false-reassurance probe.
About this page. Written and maintained by the Pymander Health team. Last updated October 19, 2026. Methodology: every factual claim on this page is limited to the published evaluations linked above, and prices and policies are stated as of the review date. Change log: October 19, 2026, first publication.
Common questions
Is the AI doctor really free?
Yes. The consult is free for everyone: no card on file, nothing billed to any plan.
Do I need to download anything?
No. It is a text message thread. If you can text, you already have everything you need.
Can I ask about my kid?
The consult is for adults 18 and over about their own health. For your child, their pediatrician is the right call - and Pymander will say so.
What can it actually tell me?
What your symptoms most likely fit, how worried to be, what to do tonight, what to watch for, and whether you need a human doctor, urgently or not. It is a medical AI system, not a licensed doctor.
What if it sounds serious?
Every consult includes a red-flag check, and a separate safety system decides when to escalate. You are told directly when something needs urgent or emergency care. In an emergency, call 911 first.
Compare Pymander head to head
See how Pymander stacks up on price and approach: Pymander vs Sesame Care, Sesame Care vs GoodRx Care, online doctor vs walk-in clinic.