Head-to-Head Architecture Comparison
When clinicians evaluate documentation systems, the bottleneck is rarely note prose. It is the last ten seconds: moving a finished draft into the correct field of the correct chart, then deciding whether the visit supports an add-on code. In primary care that decision repeats twenty times a day, so small frictions compound into hours and revenue.
Heidi Health is a capable general scribe with a strong GP consult template that numbers each problem. Its Chrome extension floats beside browser-based EHRs and copies or pushes text into the record. Merry takes a different route: a content script inside the authenticated EHR tab that reads section labels and writes into each field directly.
| Feature | Merry AI | Heidi Health |
|---|---|---|
| Real pricing | Pro Annual $54/mo ($648/yr) | Free tier; Pro $99/mo ($1,188/yr) |
| DOM injection speed | Direct write via input events; no clipboard | Copy or push into focused field |
| Chart binding | MRN and encounter header check | No page scraping; user-verified |
| Group note-splitting | Diarized per-participant notes | Not publicly documented |
| CPT G2211 capture | Drafted, with modifier 25 suppression | Not described; CPT-2025 via API |
| Attestation logging | Clinical Logic Bridge audit log | Not publicly described |
| Data retention | Audio deletion window defined in BAA | Local extension processing; DPA terms |
The table reflects public documentation at the time of review. Heidi's API lists CPT-2025 among supported coding standards, but its product pages do not describe G2211 attestation drafting or modifier 25 suppression. Verify list prices directly; both vendors adjust plans.
Pricing Math Without Metering
The arithmetic fits on a napkin. Merry AI Pro Annual is $54 per month, or $648 per clinician per year. Heidi Pro is listed at $99 per month on monthly billing, or $1,188 per year. The difference is $540 per clinician annually, about 45 percent. Heidi's free tier is useful for evaluation, but it limits features a full panel eventually needs.
For context across the market, Freed lists near $99 per month, while Nuance DAX Copilot is typically sold through enterprise contracts that commonly approach $8,000 per clinician per year. A five-physician practice on Merry spends $3,240 annually; the same group on DAX can exceed $40,000. Neither figure includes implementation time, which is where workflow friction hides.
Where the Hours Come From
Copy-paste adds friction per note. Highlight, copy, switch tabs, locate the field, paste, reformat, confirm the patient. Across twenty encounters that sequence accounts for a meaningful share of the 2.1 daily hours our pilot clinicians recovered. Direct field writes fire the EHR's own input events, so autosave and template logic behave as if you typed.
DOM Injection and Chart Safety
Merry's overlay anchors to visible section labels and ARIA roles rather than brittle element IDs, which shift whenever single-page EHR editors update. It locates Subjective, Objective, Assessment, and Plan regions and writes each section into its matching field. In iframe-based encounter frames, the overlay binds only to the note subtree.
Multi-tab charting is where errors happen. Before insertion, Merry reads the patient name, MRN, and encounter identifier from the chart header and binds the draft to that context. An attestation built for one patient cannot land in a neighboring tab, and duplicates are detected first. Heidi's extension states it does not scrape page data, a defensible privacy choice that also precludes header-based chart binding.
Implementation Directive P7
Install the Chrome overlay first, then map target EHR note fields during one supervised visit. Confirm each SOAP label resolves, test a draft on a sandbox patient, and record the mapping. Most practices finish within a single clinic session.
G2211 Capture and Modifier 25 Logic
CMS defines HCPCS add-on G2211 as visit complexity inherent to E/M care that serves as the continuing focal point for all needed services, or ongoing care for a single serious or complex condition. It attaches to office or outpatient codes 99202 through 99215 and home or residence E/M codes. Diagnosis specificity still matters; review the https://www.cms.gov/medicare/coding-billing/icd-10-codes reference when anchoring chronic conditions.
The modifier 25 rule has two dates worth keeping straight. From the code's implementation on January 1, 2024, G2211 was not payable when the base E/M carried modifier 25. Beginning January 1, 2025, CMS permits payment with modifier 25 only when the separate service is an annual wellness visit, vaccine administration, or a Medicare Part B preventive service. Merry suppresses G2211 drafting outside those carve-outs.
Implementation Directive P1
Append a focal point statement and a complexity anchor to every qualifying longitudinal visit. The statement names the ongoing relationship; the anchor names the condition, its trajectory, and coordinating clinicians. Merry drafts both in the Assessment/Plan as visible, editable text, never hidden markup, and the clinician signs.
The recovery math is conservative. At roughly $16 per G2211 claim nationally, about four qualifying Medicare visits per clinic day across 244 days yields 976 claims, or about $15,600 per physician annually. Panels with heavier chronic disease burden see more.
Audit Defense and Next Steps
Every attestation leaves a trail. Merry's Clinical Logic Bridge logs when a G2211 statement is drafted, edited, suppressed, or signed, along with the triggering signal. During a payer review, you can show why the code was claimed and why it was withheld elsewhere.
If you want to test this against your own charts, bring three recent longitudinal visits and one modifier 25 visit. Schedule a 15-Minute Workflow Audit and we will review field mapping, pricing, and G2211 eligibility side by side, the way a colleague would.
