DeepScribe
Also known as: DeepScribe AI
Specialty first ambient AI scribe with the category's top KLAS score (98.8), oncology dominance, bidirectional EHR write back, and coding built into the note.
DeepScribe is the specialty depth play in ambient documentation: an AI scribe built for complex specialty medicine rather than general primary care. Founded in 2017 in San Francisco, the company holds the highest independent rating in the category, a 98.8 KLAS Emerging Company Spotlight score with A+ marks across all six measured categories, plus 2025 KLAS Top 5 rankings for improving outcomes, patient experience, and clinician experience. Its center of gravity is oncology, with 5 million oncology visits captured a year and roughly 40 percent of US cancer care visits documented through the platform, alongside specialty tuned models for cardiology, urology, gastroenterology, orthopedics, neurology, and nephrology. An August 2025 Journal of Clinical Pathways study found meaningful improvements in diagnosis capture and note quality in oncology.
The product has evolved from scribe to what the company calls an Ambient Operating System. Bidirectional integration with Epic, athenahealth, NextGen, and eClinicalWorks reads chart context before the visit and writes structured data back into discrete fields after, and the platform pulls forward the full care journey, prior notes, labs, imaging, medications, and diagnoses, adapting it to the current day's note. AI coding builds HCC, CPT, and ICD-10 suggestions into the note itself, SmartPrep adds pre visit intelligence for oncology, and a Customization Studio learns each clinician's documentation style from day one, with the company reporting a 99.92 percent note approval rating and average chart closure of 1.6 minutes. Conversations are supported in more than 110 languages, and security includes end to end AES-256 encryption, de identified data, and multifactor authentication.
DeepScribe is enterprise sales led with no public pricing, no self serve signup, and no trial; third parties estimate roughly $350 to $500 per provider per month, with some reports near $750 quote based. The premium fits complex, longitudinal specialty care; for short low acuity visits, cheaper generalist scribes cover the job.
Vendor details
Canonical URL
https://www.deepscribe.ai
Category
Healthcare agent
Subcategory
Clinical documentation
Funding status
Independent, venture backed, founded 2017 in San Francisco. Holds the highest KLAS score in the ambient AI category (98.8 Emerging Company Spotlight with A+ in all six categories) and reports roughly 40 percent of US cancer care visits documented through its platform.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Bidirectional EHR integration with Epic, athenahealth, NextGen, and eClinicalWorks: reads chart context (problems, medications, labs, prior notes) before the encounter and writes structured data back into discrete fields after, syncing with the EHR schedule. Supports patient conversations in more than 110 languages across in person and telehealth settings from iPhone and iPad.
In practice
Your oncologists run 45 minute longitudinal consults that generalist scribes butcher. DeepScribe's oncology tuned models pull forward prior notes, labs, and imaging and document at specialty depth.
Coding happens days after the visit and revenue leaks. DeepScribe builds HCC, CPT, and ICD-10 suggestions into the note as it is written, with audit ready documentation.
Notes arrive generic and every clinician rewrites them their own way. Customization Studio learns each provider's documentation style from day one and refines from their edits.
Sources & related URLs
Capability coverage
7.0 / 14 capabilities · 50%
| Integrations & Tool CallingBidirectional integration with Epic, athenahealth, NextGen, and eClinicalWorks: reads chart context pre visit and writes structured data into discrete fields post visit with schedule sync, DeepScribe docs 2026-07-06 | Full |
|---|---|
| Workflow OrchestrationSmartPrep pre visit intelligence, note generation, and AI coding chained across the encounter; fixed function pipeline, DeepScribe docs 2026-07-06 | Partial |
| Knowledge Grounding & RAGPulls forward the full care journey (prior notes, labs, imaging, medications, diagnoses) into the current note; scoped to the patient record, DeepScribe docs 2026-07-06 | Partial |
| Human Oversight & GuardrailsClinician review and sign off on every note; the earlier human QA review layer is no longer promoted as a core differentiator as of 2026, DeepScribe docs 2026-07-06 | Partial |
| Security, Identity & GovernanceHIPAA with end to end AES-256 encryption, de identified patient data, multifactor authentication, and access controls, DeepScribe docs 2026-07-06 | Full |
| Observability & AuditabilityAudit ready coding documentation and compliance support; no per sentence note provenance tooling, DeepScribe docs 2026-07-06 | Partial |
| Memory & State PersistenceCustomization Studio is a first class personalization system: learns each clinician's documentation style from day one and refines from edits, with a reported 99.92 percent note approval rating, DeepScribe docs 2026-07-06 | Full |
| Deployment & Data ResidencySaaS only with US processing; no self host or residency options documented, DeepScribe docs 2026-07-06 | Partial |
| Prebuilt Agents, Templates & PacksSpecialty tuned models are the flagship: oncology, cardiology, urology, gastroenterology, orthopedics, neurology, and nephrology, with roughly 40 percent of US cancer care visits documented, DeepScribe docs 2026-07-06 | Full |
| Triggers & Channel CoverageAmbient capture from iPhone and iPad across in person and telehealth settings in 110 plus languages; no event trigger system, DeepScribe docs 2026-07-06 | Partial |
| Model Flexibility & RoutingProprietary specialty tuned models; no customer facing model choice or routing, DeepScribe docs 2026-07-06 | Unable to verify |
| APIs, SDKs & MCP ExtensibilityNo public developer API or SDK; integration runs through native EHR connections only, DeepScribe docs 2026-07-06 | Unable to verify |
| Testing, Debugging & OptimizationPublished outcome studies but no customer facing testing or evaluation tooling documented, DeepScribe docs 2026-07-06 | Unable to verify |
| Browser & Computer UseNo browser or computer use capability, DeepScribe docs 2026-07-06 | Unable to verify |
Pricing
Contact sales; third parties estimate $350 to $500 per provider per month, some reports near $750
seats (per provider per month)
Included quota
Enterprise license per provider including specialty tuned models, bidirectional EHR integration, AI coding, pre visit intelligence, and the Customization Studio.
What is public
Nothing numeric from the vendor; only third party estimates with a wide spread.
Billing mechanics
Per provider enterprise subscriptions negotiated per organization on size, specialty mix, and EHR integration depth, with pilots preceding rollouts. No individual plan, no self serve path.
Cost watchouts
Estimates span a wide range ($350 to $750 plus per provider per month) depending on source, organization size, and integration depth, and the enterprise process means no way to test in your own clinical environment before procurement. Hard to justify for short low acuity visit mixes.
Variable cost rationale
Flat per provider enterprise license; organization size and integration depth move the negotiated rate, not usage.
Additional watchouts
The most premium priced specialist scribe by most estimates: strong fit for complex longitudinal specialty care, hard to justify for simple primary care workflows where generalist scribes at a tenth of the price cover the job.
Overage / add-ons
No usage metering documented; flat per provider enterprise licensing.
Sales call required
Yes — required for paid access
Free / trial
No self serve trial; organizational pilots through sales
Lowest paid plan
None public; enterprise contract only
Commercial notes
Independent, founded 2017 in San Francisco. Highest KLAS score in ambient AI (98.8, A+ across all six categories), roughly 40 percent of US cancer care visits documented, 5 million oncology visits captured per year.
Key ambiguities
Third party estimates disagree materially: $350 to $500 per provider per month in one recent review versus roughly $750 quote based in another. Budget from a negotiated quote.
Related vendors
- Abridge — Enterprise ambient AI documentation platform, Best in KLAS 2025 and…
- Ambience Healthcare — Enterprise AI documentation suite pairing the AutoScribe ambient…
- Commure — General Catalyst backed roll up (Athelas, Augmedix, Memora) selling…
- Freed — Self serve ambient AI scribe with public pricing from $39 per month,…
- Glass Health — Clinical AI platform pairing ambient scribing with evidence grounded…
- Heidi — Global ambient AI scribe supporting 2