Medical scribe software turns an encounter into a draft note for clinician review, then hands the reviewed result to the record. A clinic has four separate decisions: what comes out, who reviews it, how long the vendor keeps data, and how the note reaches the chart. Account ownership, offboarding, and seat pricing come next. Don’t assume the word clinic guarantees any of them.
Key takeaways
- Separate output quality from the operating workflow around it. A polished draft can still fail on review, retention, or chart handoff.
- Make the vendor demonstrate account ownership, offboarding, export, and price for every clinician type you plan to activate.
- Price the exact tier and active seats you will use. Public rates and quote-only contracts aren’t comparable until limits and renewal terms are written down.
- Start with a one-week failure log on real visits. Expand only after clinicians can repeat the review and handoff without guesswork.
- A solo-only tool or a human scribe service is the better fit in specific cases. We name them below.
| Capability | Question to prove | Evidence to collect |
|---|---|---|
| Output | What the tool drafts and what it never decides | A marked-up note from a difficult real visit |
| Review | Who checks, edits, and signs | Minutes to sign plus a log of missing or invented facts |
| Retention | What happens to audio, transcript, note, and account data | Written retention, deletion, and cancellation terms |
| Handoff | Copy, file export, structured export, extension, or native write-back | One completed transfer into the clinic record |
| Clinic operations | Account owner, roles, offboarding, active-seat billing | A live admin demonstration and a sample invoice |
Output
Review
Retention
Handoff
Clinic operations
What does “clinic-scale” actually change about scribe software?
Not the medical-content standard. CMS still expects a complete, legible record with the reason for the encounter, relevant findings, assessment, plan, date, and identifiable author (CMS E/M guidance). What changes at clinic scale is the number of people who must repeat the capture, review, handoff, and offboarding process without inventing private workarounds.
Picture a five-provider family clinic on a 14-patient Tuesday. Each provider walks out of every visit with a chart to finish. The documentation problem is the same one a solo doctor has, times five, and it shows up in the same place: the AMA, reporting a 2022 JAMA Network Open study, found primary-care physicians spent about 36 minutes on the EHR per 30-minute visit, roughly 6 of those minutes as after-hours “pajama time” (Rotenstein et al., JAMA Network Open, 2022, via AMA). Five providers, five evenings.
So the draft is only one part of the buy. Who owns the account? What happens when a provider joins in March and leaves in September? Where do encounter artifacts live, and what gets deleted? How does a reviewed note reach the chart? A vendor may handle these well, poorly, or not at all. The clinic has to collect the answer instead of filling the blank with hope.
What should a clinic require from scribe software?
Start with four evidence packets. A clinic shouldn’t sign without them.
- Output and review evidence. Mark up ten difficult notes. Record omissions, invented details, and minutes to sign. A clean demo note proves almost nothing.
- Retention and deletion terms. Ask separately about audio, transcript, draft, signed note, support logs, and backups. Get the cancellation behavior in writing.
- Handoff proof. Move one reviewed note into the real chart. Patient Square describes its output as “EHR-ready export (PDF · HL7 · FHIR).” That is an export claim, not certified interoperability or native write-back.
- Clinic controls and cost. Ask the vendor to show account ownership, roles, offboarding, active seats, one sample invoice, and renewal terms. If a capability isn’t shown, treat it as absent from the decision.
Our opinion: the failure log and handoff test expose more than any feature grid. If the vendor can’t answer these four packets, the clinic isn’t ready to buy it.
How much does clinic scribe software cost across N providers?
Many ambient AI scribes price per clinician per month; others offer a free eligible tier or a sales quote. For a clinic, the unit, usage limit, add-ons, renewal term, and meaning of an active seat all matter.
Patient Square publishes $79 per clinician per month on the annual Assist plan; the Copilot plan adds a bundled AI EHR and WhatsApp messaging at $119, and Autopilot is $199 with an AI receptionist and follow-ups (Patient Square US pricing). Freed publishes its own tiers (Freed pricing). Don’t turn those two examples into a market average. Put every finalist’s live rate and contract terms into the same worksheet.
Two things move the clinic number:
- Active seats, not headcount. A part-time provider who documents 20 notes a month may still cost a full seat. The contract controls. Model your bill on seats you will actually turn on, and ask whether a low-volume clinician can sit on a cheaper tier.
- What’s bundled changes the comparison. A $79 scribe-only seat and a $119 seat that includes an EHR aren’t the same purchase. When you compare vendors, line up what each seat contains, not just the number.
For the per-clinician version of this math and a longer pricing walk-through, the scribe pricing breakdown covers the solo case in detail, and the group-practice per-seat guide works the multi-provider contract questions.
How do you roll scribe software out across a whole clinic?
Stage it. Run a one-week pilot first, study the failures, then decide whether another specialty or clinician should join. Buying seats for everyone on day one is how a rollout stalls quietly.
Here’s the order we’d run it in a five-to-fifteen-provider clinic. The full rollout may be quick or may take months; integration, contracting, and the failure rate decide.
- Week one: pilot 3 to 5 providers on real visits. Not a demo, real Tuesdays. The pilot finds the note-quality quirks and the specialties where the draft needs more editing. A sales demo won’t.
- Standardize one note structure. Pick a shared SOAP order so a covering provider can open a colleague’s chart at 7pm and find the plan where they expect it. Let clinicians edit within it; don’t let every provider invent a private format.
- Set ownership and billing before you scale. Decide who owns the vendor account, how active seats are counted, who approves access, and how a departing clinician is removed. Make the vendor demonstrate each control it actually provides.
- Add the rest in a staged wave. Bring on another specialty or small group only after the first group can capture, review, and hand off a note without private instructions.
The small-clinic rollout plan walks a two-week version of this step by step if you want the day-by-day version.
When is a solo-only tool or a human scribe the better fit?
Two honest cases where clinic-scale AI software is the wrong buy.
A genuinely solo practice. If you’re one clinician with no plans to add providers, the clinic admin layer is overhead you’ll never use. A solo-tier scribe with a single account is cleaner and often cheaper. The solo pricing breakdown is the right starting point; skip the multi-provider machinery entirely.
A clinic with unusually complex visits, or a provider who wants a person. A human scribe service puts a trained person on the documentation, live, and some clinicians strongly prefer that for high-complexity encounters. The tradeoff is cost and turnover: a human scribe is a recurring salary or contract, and scribes leave, so you re-hire and re-train. The AI-versus-human cost and turnover math lays out both sides. For most multi-provider clinics chasing steady per-seat cost with no staffing churn, AI software wins. For a few, the human fit is real, and we’d rather you buy the right thing.
Where Patient Square fits for a clinic
The AI Medical Scribe by Patient Square is one module of Practice Copilot. It captures the visit and returns a structured SOAP note, ICD-10 suggestions, and a prescription draft for clinician review. Scheduling starts on Assist; the Copilot tier adds the bundled AI Copilot EHR and WhatsApp messaging. These are plan claims, not proof of a particular clinic’s admin or EHR write-back needs.
The public rate is per clinician: $79 per month on the annual Assist plan, or $119 on Copilot. Every plan includes unlimited visits and notes, no per-note metering, and EHR-ready export (PDF · HL7 · FHIR). The site doesn’t claim native write-back to every third-party EHR. A clinic should still ask us to demonstrate account ownership, offboarding, and the exact handoff it intends to use.
If you’re evaluating for a multi-provider clinic,
Bottom line on clinic scribe software
The clinic decision has four parts: output, review, retention, and handoff. Prove those first. Then verify account ownership, offboarding, price, and renewal terms. A one-week failure log gives you more useful evidence than a big launch and a stack of seats.
If you’re truly solo, buy the solo tool. If your visits are unusually complex and you want a person, price a human scribe honestly. For clinics that want to test steady per-provider cost and less time spent writing notes,
Frequently asked questions
The FAQ above covers the recurring ones: what clinic scribe software is, what it costs across providers, the four things to require, the AI-versus-human call, EHR export, rollout time, and non-physician use. For the full plan detail, the pricing page has every tier.
How Patient Square fits your existing record system
In the US, an established practice can keep its current EHR as the system of record while Practice Copilot supports documentation and shipped practice workflows. A new practice can instead use Patient Square EHR as its system of record from the Copilot tier. Neither path promises automatic filing or write-back into a named third-party EHR.