Sign Up
Login
Try for Free
Login
This page records how the figures in our comparison tables are produced: what was measured, where it came from, when it was read, and what it does not prove. Twenty-two products, covering ambient AI medical scribes and the healthcare and general AI assistants used alongside them. BastionGPT appears in both groups because it is both.
A rating is the mean of a product’s scores across the review sites where it holds a published rating, read 16 August 2026. Each cell below shows that site’s own score and review count, and links to the page it was read from, so every average can be checked against its parts. The rules are set out beneath the tables.
| Product | Averageof 5 | Apple App Store | Capterra | G2 | GetApp | Google Play | Software Advice | Trustpilot | TrustRadius | |
|---|---|---|---|---|---|---|---|---|---|---|
| Abridge | 4.20 | 4.245 | listedno rating | 4.54 | · | · | 3.921 | · | listedno rating | · |
| Ambience Healthcare | — | · | · | listedno rating | · | · | · | · | · | · |
| BastionGPT | 4.84 | · | 4.84 | 4.84 | 4.84 | 55 | · | 4.84 | · | · |
| Commure | 4.41 | 4.9155 | · | 4.416 | · | · | 3.9212 | · | listedno rating | · |
| DeepCura | 4.75 | · | 4.815 | 4.613 | 4.815 | · | · | 4.815 | · | · |
| DeepScribe | 4.20 | 439 | 4.38 | 4.126 | 4.38 | · | · | 4.38 | · | listedno rating |
| Dragon Copilot | 4.20 | 4.114 | · | 4.326 | · | · | 4.813 | · | · | 7.2/1035 |
| Freed AI | 4.80 | 4.8515 | 52 | 54 | · | · | 4.4~100 | · | · | · |
| Heidi Health | 4.15 | 4.72.4K | listedno rating | 53 | listedno rating | · | 3.7~350 | listedno rating | 3.2490 | · |
| Mentalyc | 4.56 | · | 4.561 | 4.85 | 4.561 | · | · | 4.561 | 4.570 | listedno rating |
| Nabla | 3.90 | 3.98 | · | listedno rating | · | · | · | · | listedno rating | · |
| Suki AI | 2.73 | 4.165 | 21 | 2.51 | 21 | · | 3.833 | 21 | · | listedno rating |
| Tali | 4.60 | · | 4.621 | · | · | · | · | · | · | · |
| Twofold Health | 4.93 | 4.852 | 51 | 53 | · | · | listedno rating | · | listedno rating | · |
| Upheal | 4.43 | 3.712 | listedno rating | 52 | · | · | listedno rating | · | 4.674 | · |
| Product | Averageof 5 | Apple App Store | Capterra | G2 | GetApp | Google Play | Software Advice | Trustpilot | TrustRadius | |
|---|---|---|---|---|---|---|---|---|---|---|
| BastionGPT | 4.84 | · | 4.84 | 4.84 | 4.84 | 55 | · | 4.84 | · | · |
| ChatGPT | 3.77 | · | 4.4380 | 4.62,754 | · | · | · | · | 1.63,280 | 9/10640 |
| Claude | 3.75 | · | 4.346 | 4.6395 | · | · | · | · | 1.51,672 | 9.2/10236 |
| CompliantChatGPT | — | · | · | · | · | · | · | · | · | · |
| Doximity GPT | — | network4.8 · 195K | · | network4.6 · 9 | · | · | network4.8 | · | network3.0 · 5 | · |
| Hathr AI | — | · | listedno rating | listedno rating | listedno rating | · | · | listedno rating | · | · |
| Microsoft Copilot | 3.93 | · | 4.527 | 4.4358 | 4.527 | · | · | 4.527 | 1.6392 | 8.1/10187 |
| OpenEvidence | 4.17 | 4.911K | · | 52 | · | · | 4.94.36K | · | 1.926 | · |
listed — a profile exists with no published rating. Not a zero; excluded from the mean. network — the figure belongs to the parent Doximity network app, not to Doximity GPT; shown for provenance, excluded from the mean. A dot means no profile exists on that site.
A security grade is an MDN HTTP Observatory scan of the host where clinicians sign in, run 26 July 2026 and re-checked 16 August 2026. Every grade links to its own scan, so you can re-run it and see today’s result rather than ours. What the grade does and does not cover is set out beneath the tables.
| Product | Marketing site | Sign-in hostwhere you log in |
|---|---|---|
| Abridge | D30abridge.com | A+105notes.abridge.com |
| Ambience Healthcare | C50ambiencehealthcare.com | B+80app.ambiencehealthcare.com |
| BastionGPT | B75bastiongpt.com | B+80secure.bastiongpt.com |
| Commure | B70commure.com | C50scribe.commure.com |
| DeepCura | F15deepcura.com | A+110app.deepcura.com |
| DeepScribe | B70deepscribe.ai | D−25portal.deepscribe.ai |
| Dragon Copilot | D−25microsoft.com | C50www.copilot.us.dragon.com |
| Freed AI | D−25getfreed.ai | B+80secure.getfreed.ai |
| Heidi Health | C50heidihealth.com | C−45scribe.heidihealth.com |
| Mentalyc | C−45mentalyc.com | D30app.mentalyc.com |
| Nabla | C−45nabla.com | C50copilot.nabla.com |
| Suki AI | B+80suki.ai | B70app.suki.ai |
| Tali | F0tali.ai | F0app.tali.ai |
| Twofold Health | C50trytwofold.com | C50app.trytwofold.com |
| Upheal | B70upheal.io | A+100app.upheal.io |
| Product | Marketing site | Sign-in hostwhere you log in |
|---|---|---|
| BastionGPT | B75bastiongpt.com | B+80secure.bastiongpt.com |
| ChatGPT | B75openai.com | B75chatgpt.com |
| Claude | B75claude.com | C+60claude.ai |
| CompliantChatGPT | C50compliantchatgpt.com | B75app.compliantchatgpt.com |
| Doximity GPT | B70doximity.com | Same host |
| Hathr AI | B70hathr.ai | D30app.hathr.ai |
| Microsoft Copilot | D−25microsoft.com | A+115m365.cloud.microsoft |
| OpenEvidence | C−45openevidence.com | Same host |
Marketing sites carry analytics, advertising and chat tags that require loosening exactly the headers Observatory tests, so grading them measures a tag manager rather than a product. The host behind the login is where credentials and patient information are typed. Where a vendor does not publish its application hostname, we resolved the conventional one and confirmed it serves a real sign-in application before scoring it. The host scored is printed under every grade.
HTTP response headers only: HTTP Strict Transport Security, Content Security Policy, X-Frame-Options, X-Content-Type-Options, Referrer-Policy, cookie flags and subresource integrity. Scores run 0–100, with bonus points above 100.
It earns a column for one reason: it is among the very few security signals a buyer can verify independently, in a browser, without taking any vendor’s word for it — including ours. Score reports from BitSight and SecurityScorecard could not be included here because of licensing restrictions.
Prices were read from each vendor’s public pricing page on 16 August 2026, and every price in the comparison tables links to the page it was read from.
Five of the ten platforms in the scribe comparison publish no price at all — Abridge, Ambience Healthcare, DeepScribe, Dragon Copilot and Suki AI each route buyers to a sales conversation. Those are marked “Custom”, and any figure beside them is a third-party estimate rather than a quote. Implementation fees, reseller markups, seat minimums and annual prepayment terms sit on top of whatever is eventually quoted and are not visible to us. Prices are per user per month at the entry tier, before annual discounts.
If a figure here is wrong, tell us and we will correct it. That includes a price we misread, a review profile we missed, an application hostname we resolved incorrectly, or a capability we recorded as unpublished that is in fact documented.