Dragon Copilot Licensing

What Dragon Copilot is (and what it replaces)

Dragon Copilot is Microsoft's unified clinical AI assistant. It combines three predecessor capabilities into one commercial product: Dragon Medical One (real-time ambient dictation into the EHR — the historic Nuance speech-recognition engine), DAX Copilot (ambient encounter documentation that listens to a patient-provider conversation and drafts the clinical note), and EHR-grounded clinical assistance (chart summarization, order navigation, intelligent search across the patient record).

Commercially, Dragon Copilot is the forward-going SKU. Microsoft is migrating existing Dragon Medical One and DAX Copilot customers onto the unified product, and net-new sales go directly into Dragon Copilot. Customers on prior-product paper have a discrete negotiation event at conversion — the migration price is not a foregone conclusion, and prior-term unit-economics protections are recoverable if pushed at the right moment.

The Dragon Copilot SKU map

Dragon Copilot is sold in three primary commercial constructs plus specialty add-ons:

• Dragon Copilot Physician — named-user per-physician subscription, unlimited use within the entitlement, base ambient dictation + ambient documentation. The default SKU for full-time clinicians.
• Dragon Copilot Physician Flex — consumption-based SKU billed per encounter (or per minute, depending on contract construct). Intended for residents, locums, fellows, occasional users, and ambulatory-only providers below a threshold of monthly usage.
• Dragon Copilot Specialty — per-physician add-on modules with specialty-trained models for cardiology, radiology, pathology, surgery, and a growing list of specialties. The default Dragon Copilot Physician model is generalist; the specialty add-on materially improves note quality and acceptance for that specialty.

The trap most health systems fall into is buying Dragon Copilot Physician seats for all clinicians without modeling the Flex blend. A typical 1,200-physician health system with 280 part-time/locum/resident providers can shift those 280 to Flex at 30-50% lower total cost than blanket per-physician seats, with no service degradation for those users.

Physician Flex: when consumption beats per-seat

Dragon Copilot Physician Flex is the SKU Microsoft sales motions sometimes leave out of the opening proposal. The breakeven calculation is the per-encounter or per-minute Flex price multiplied by typical monthly usage, compared to the fixed per-physician monthly subscription. Below the breakeven volume — typically 60-90 encounters per month for ambient documentation use cases — Flex is materially cheaper.

The user populations that consistently belong on Flex in our health-system engagements: graduate medical education residents (variable rotations, low standalone documentation volume), locum tenens physicians on short contracts, telehealth-only providers for telecom-only workflows, ambulatory specialists with low office days, and any provider with documented monthly encounter counts below the breakeven threshold.

Specialty modules and their per-physician economics

The Specialty add-on is per-physician per-month on top of the base Physician seat. The clinical case is real — a cardiology-tuned ambient documentation model produces materially better notes for cardiology encounters than the generalist model — but the financial case is per-specialty. Five disciplines see the strongest return on the Specialty add-on:

1. Radiology — historic Dragon strength; specialty model is the most mature.
2. Cardiology — high-volume documentation, specialty vocabulary, dense order patterns.
3. Pathology — structured reporting, specialty vocabulary.
4. Surgery — operative note structure, intra-op dictation patterns.
5. Emergency Medicine — high encounter velocity, structured documentation expectations.

For other specialties, the generalist Physician model is usually adequate and the Specialty add-on is incremental spend without proportional benefit. Negotiate the Specialty add-on per-discipline, not as a blanket organization-wide commit.

EHR integration: Epic, Cerner, MEDITECH, athena

Dragon Copilot's commercial value depends on how cleanly it integrates with your EHR. Integration depth and maturity vary by EHR vendor.

Validate at module level for your specialty mix. The Microsoft demo will use a best-case EHR pairing; your environment will hit edge cases. Negotiate acceptance criteria into the contract, particularly for specialty-module performance against your EHR's note templates.

The DAX Copilot → Dragon Copilot transition

If you are on a DAX Copilot or Dragon Medical One contract, your conversion to Dragon Copilot is a discrete commercial negotiation, not a routine renewal. Microsoft will typically open with the new SKU at list (or near-list) for the converted seats, plus a "transition bridge" of 6-12 months of prior-term pricing. That construction is the opener, not the answer.

Three things to push at conversion:

1. Multi-year price protection. Lock the unit price for the full new term, not just the bridge period.
2. Like-for-like entitlement. If DAX Copilot's prior pricing included unlimited use and the new Dragon Copilot Physician SKU has similar entitlement, the unit conversion should be at parity, not at a premium.
3. Flex availability. Surface Flex at conversion — it is rarely in the opener — and shift eligible providers to it as part of the conversion structure.

See our Copilot Advisory service for the migration playbook.

Contract structure and discount tiers

Standard Dragon Copilot contracts are 3-year subscriptions with annual true-up. Discount tiers step at 100, 500, and 1,500 physician seats; large IDNs above 5,000 seats negotiate bespoke pricing that does not follow the published tier structure. Five-year terms with prepayment commitments unlock additional 8-15 points of concession on the right deal construction.

The contractual items to negotiate beyond unit price:

• True-up timing. Annual is typical; quarterly is rarely necessary and exposes you to ratchet-up risk if utilization fluctuates.
• Down-true rights. Provider churn happens — negotiate the right to reduce seat count at anniversary, not just add to it.
• Specialty module termination. If a specialty add-on does not deliver clinical value, the right to remove it without penalty at anniversary should be explicit.
• BAA and HIPAA terms. Standard, but validate the current language matches your privacy office's requirements.
• Implementation services. Negotiate as fixed-fee or capped, not T&M. Microsoft's implementation arm and partner network both work this market.

Clinical, security, and HIPAA governance

Dragon Copilot is delivered under Microsoft's healthcare cloud commitments: BAA-covered, HIPAA-compliant data handling, audit logging, and data-residency options for US, EU, UK, and select APAC regions. The default deployment is multi-tenant SaaS; dedicated-instance and sovereign cloud options exist for specific regulatory regimes (US Government Community Cloud, UK NHS, EU sovereign).

The clinical governance load is on the provider organization, not Microsoft. Note-quality review, hallucination detection, encounter recording consent, and the clinical-decision-support boundary are health-system responsibilities. Build the governance framework before the rollout — Microsoft's product team has materially improved guardrails since the 2024 launch, but the organizational acceptance criteria still belong to clinical leadership.

Negotiation levers for health systems

Five levers consistently move Dragon Copilot economics:

1. Right-sized Physician/Flex split. Audit your provider base by usage and shift eligible providers to Flex. 15-30% of provider populations belong on Flex; the opening proposal will rarely include the breakdown.

2. Discipline-level Specialty add-on. Negotiate Specialty per discipline based on clinical case, not as a blanket commit.

3. Multi-entity consolidation. Health systems with multiple legal entities should contract through the largest single entity to maximize tier discount.

4. EHR integration acceptance criteria. Bake measurable acceptance criteria into the contract — specialty-module accuracy thresholds against your EHR, specific EHR-version support guarantees, integration roadmap milestones.

5. Microsoft Cloud for Healthcare bundling. If you are also negotiating Microsoft Cloud for Healthcare components (FHIR services, MedTech connectors, Power Platform for healthcare), bundle the Dragon Copilot conversation into the larger commercial envelope. Our EA Negotiation service runs this combined construct.

Major 2026 changes affecting Dragon Copilot licensing

Three named changes:

1. Unified product maturity. Dragon Copilot replaced Dragon Medical One and DAX Copilot in commercial sales through 2025. By 2026, the migration cohort is the active negotiation surface — health systems on prior paper are converting in waves, and Microsoft's conversion economics tighten as migration completion approaches.

2. Specialty module expansion. Three additional specialty modules entered general availability in 2026 (pediatrics, obstetrics, internal medicine). New specialty availability is an opportunity to renegotiate the Specialty bundle at point of expansion.

3. Agent 365 governance overlay. Dragon Copilot does not require Agent 365, but health systems building AI governance frameworks for downstream agents (revenue cycle, prior authorization, clinical pathways) are increasingly attaching Agent 365 to govern the broader portfolio. See our Copilot Licensing Guide for the cross-portfolio view.