//
Copenhagen healthcare AI infrastructure with sovereign EU hosting for clinical documentation
Review by EuropeanStack EditorialUpdated Verified
Corti's sovereign EU cloud claim is unusually well-earned: a decade of high-stakes clinical deployment, starting in emergency dispatch, backing a genuinely European infrastructure build rather than a rebadged hyperscaler region. Combined with published, usage-based pricing rare in healthcare AI, it is a strong choice for developers and health-tech vendors building on top of clinical AI rather than buying a finished app. The trade-offs are the ones you would expect from developer infrastructure — real engineering effort required, a narrower named-integration list than EHR-native rivals, and a total-funding figure that varies by source. For the audience it targets, those are reasonable costs for a genuinely differentiated EU data-sovereignty story.
Corti is a Copenhagen-based healthcare AI infrastructure company providing an API for medical speech-to-text, clinical documentation, medical coding, and agentic workflows. Founded in 2016 by Andreas Cleve and Lars Maaloe, it started by detecting cardiac arrest on emergency calls and has since expanded into a sovereign EU cloud for healthcare AI, processing more than a million clinical interactions a week.
Headquarters
Copenhagen, Denmark
Founded
2016
Pricing
EU Data Hosting
Yes
Employees
51-200
Pay-as-you-go
Pay-as-you-go
Contact Sales
Contact Sales
Billing: pay-as-you-go, monthly
Corti calls its 2025 infrastructure launch Europe's first sovereign healthcare AI cloud, and the claim holds up better than most vendor superlatives. Rather than routing clinical AI workloads through a US hyperscaler with an EU region flag, Corti built a dedicated European cluster instead. There is no external hyperscaler dependency — every prompt and workload stays on European infrastructure for customers who need it.
That bet on sovereignty did not start as a compliance feature. Corti was founded in Copenhagen in 2016 by Andreas Cleve and machine-learning researcher Lars Maaloe. It began by detecting cardiac arrest on emergency calls, flagging critical illness from speech patterns, dialect, and background noise during live calls. The system worked faster than a human dispatcher could, and Copenhagen's own emergency services deployed the model on the city's 112 line. That deep, high-stakes clinical use case shaped a company that treats reliability and data governance as core product requirements, not marketing add-ons applied after the fact.
Today Corti has expanded well beyond emergency dispatch into a broader healthcare AI infrastructure platform. That platform spans medical speech-to-text, clinical documentation generation, medical coding, and an agentic framework for building clinical workflow agents, unified under what the company calls Symphony. It processes more than a million clinical interactions weekly, with customers including the NHS, Ramsay Sante, and Trifork. Corti sits in EuropeanStack's Healthcare AI category alongside Paris-based Nabla, a useful comparison since the two take different approaches to the same underlying problem — Corti as developer infrastructure, Nabla as a clinician-facing scribe.
In the wider healthcare AI market, Corti also sits opposite US ambient-documentation vendors such as Abridge, though the comparison is imperfect. Abridge sells a finished clinical documentation product to health systems directly. Corti instead sells the underlying API layer, which a health-tech vendor, an EHR company, or an internal engineering team can build a similar product on top of. For a European health system evaluating both categories, that distinction — buy a finished product versus build on infrastructure — matters as much as any feature comparison.
Corti's speech-to-text API is tuned specifically for medical audio — clinical terminology, multiple speakers, and the background noise of a real consultation or emergency call — rather than a general-purpose transcription model repurposed for healthcare. Pricing starts at $0.0065 per minute of audio.
The coding API returns structured ICD-10 and CPT predictions along with supporting evidence and a complete audit trail, useful for compliance teams that need to justify a coding decision rather than trust a black-box output. This positions Corti as infrastructure for other software vendors — EHR companies, virtual care platforms, practice management tools. Those vendors build coding features on top of the API, rather than a single practice using it as a direct coding tool. A mis-transcribed term or a miscoded claim can still slip through any speech-to-text or coding pipeline, so the audit trail is there for a clinician or biller to check, not to replace that check.
Corti's agentic framework lets developers build clinical AI agents that reason, plan, and act across workflows. More than 20 pre-built agents back the framework, including an ICD-10 Navigator, a Compliance Guardrail, Medication Reconciliation, and Denial Appeals agents. This is a meaningfully more developer-oriented feature set than most ambient scribe competitors offer, closer in spirit to an API platform like Dust than to a packaged clinical app.
The sovereign cluster, launched in 2025, is Corti's clearest differentiator. It is designed to operate in the most restrictive EU jurisdictions, with GDPR compliance built in by default and the flexibility to exit any cloud provider if regulatory or geopolitical needs demand it. Unlike vendors that rely entirely on hyperscaler-managed regions, Corti built the underlying architecture to run across any cloud or on-premises data centre.
Every API — speech-to-text, coding, text generation, the S1 model family — draws from one shared credit wallet, rather than separate billing relationships per product. Credits are valid for one year, which rewards steady usage but can penalise low-volume or seasonal customers who let credits lapse unused.
Corti ships a JavaScript SDK, a .NET SDK, and a dedicated Agent SDK, covering the two most common stacks among health-tech vendors and EHR integration teams. Beyond documentation, the Acceleration Pack and Corti Advance tiers add real developer support hours rather than leaving integration entirely to self-serve docs. That support matters for a smaller engineering team building its first clinical AI feature without in-house healthcare AI expertise.
Corti's pricing is the rare bright spot in a category dominated by contact-sales opacity. New accounts get $50 in free credits with no upfront commitment. From there, speech-to-text runs from $0.0065 per minute, and text and coding APIs run from $4 per million input tokens and $16 per million output tokens. The newer Corti S1 model family starts even lower, at $1 per million input tokens. An Acceleration Pack bundles 1,000 monthly credits with five hours of developer support. Corti Advance adds embedded engineering help on a rolling commitment, and Enterprise covers custom bundles and sovereign cloud deployment.
This transparency is unusual against Nabla's contact-sales pricing and most of the ambient-scribe category generally. It suits developers who want to estimate cost before committing. That transparency comes with a trade-off, though: Corti is fundamentally a build-on-top platform rather than a finished clinician-facing product, so non-technical buyers need an engineering team to actually use the API.
Corti is a Danish company, Corti ApS, operating under GDPR and Danish law by default. Its sovereign EU cloud option is a genuine architectural commitment, not a checkbox — the company built dedicated European infrastructure because hyperscaler-managed EU regions were not sufficient assurance for the most restrictive healthcare customers.
The company states NIS2 alignment and EU AI Act alignment, both relevant to any healthcare software vendor operating in Europe under tightening regulatory scrutiny. Corti also holds SOC 2 Type II certification and describes itself as HIPAA-ready for customers operating in the US market. Local ownership sits with the founders' Danish holding companies and investors including Prosus Ventures, Atomico, and Danish state investment fund EIFO. Small local sales subsidiaries in the US and UK contract EEA customers back to the Danish parent entity.
EHR vendors and health-tech platforms building their own documentation or coding features get infrastructure to build on, not a competing finished product, avoiding the buy-versus-build tension that comes with a clinician-facing tool like Nabla.
European health systems with strict data-residency mandates get a real sovereign-cloud option, not just an EU-region toggle inside a US hyperscaler account.
Startups and small teams needing predictable costs benefit from Corti's published per-minute and per-token pricing, letting them model API spend before signing anything.
Solo clinicians wanting a plug-and-play scribe app should look at a clinician-facing product instead — Corti's API-first design assumes an engineering team sits between the platform and the end user.
Public-safety and emergency-response organisations get a vendor with genuine domain history in the space, since Corti's earliest deployments were on live emergency call lines rather than in outpatient clinics.
Corti's sovereign EU cloud claim is unusually well-earned: a decade of high-stakes clinical deployment, starting in emergency dispatch, backing a genuinely European infrastructure build rather than a rebadged hyperscaler region. Combined with published, usage-based pricing rare in healthcare AI, it is a strong choice for developers and health-tech vendors building on top of clinical AI rather than buying a finished app. The trade-offs are the ones you would expect from developer infrastructure — real engineering effort required, a narrower named-integration list than EHR-native rivals, and a total-funding figure that varies by source. For the audience it targets, those are reasonable costs for a genuinely differentiated EU data-sovereignty story.
Yes. Corti is a Danish company (Corti ApS) and is GDPR compliant, with a sovereign EU cloud option that keeps prompts and workloads entirely on European infrastructure. The company is also NIS2-aligned and states EU AI Act alignment.
Corti publishes usage-based pricing: speech-to-text starts at $0.0065 per minute, and text and coding APIs start at $4 per million input tokens. New accounts get $50 in free credits, with Acceleration Pack, Advance, and Enterprise tiers for higher usage.
Corti is an API platform for healthcare AI: medical speech-to-text, clinical documentation generation, ICD-10 and CPT coding, and an agentic framework for building clinical workflow agents, unified under its Symphony platform.
The platform offers a sovereign EU cloud cluster with no dependency on external hyperscalers, designed to keep every prompt and workload on European infrastructure for customers in restrictive EU jurisdictions.
Corti raised $60M in a Series B round in September 2023, co-led by Prosus Ventures and Atomico. Third-party trackers report total funding in the $94.5M to $100M range, though Corti has not published one official lifetime total.