API directory · General speech-to-text service
Azure AI Speech API
General-purpose speech-to-text with fast, batch and real-time modes and customisable vocabulary. There is no medical-specific SKU — clinical terminology is a customisation exercise.
Short answer
What does the Azure AI Speech API give a developer?
General-purpose speech-to-text with fast, batch and real-time modes and customisable vocabulary. There is no medical-specific SKU — clinical terminology is a customisation exercise.
Base URL: https://<region>.api.cognitive.microsoft.com · realtime: wss://<region>.stt.speech.microsoft.com
Auth: Ocp-Apim-Subscription-Key header, or Entra ID bearer token
Cite as: Azure AI Speech API listing, Compare Healthcare API, last reviewed 2026-09-01.
Endpoints
Copied from the vendor's own reference. Where a path is described rather than published, the description is what the vendor states.
| Method | Endpoint | Mode | Purpose |
|---|---|---|---|
| POST | /speechtotext/transcriptions:transcribe?api-version=2025-10-15 | Batch | Fast transcription: synchronous, faster-than-real-time for audio under 2 hours / 250 MB. |
| POST | /speechtotext/transcriptions (batch REST) | Batch | Create an asynchronous batch transcription job. |
| WS | Real-time speech-to-text (Speech SDK) | Streaming | Streaming recognition with diarisation and custom models. |
Published rate limits
Numbers the vendor states publicly. Anything unpublished should be made contractual before you plan capacity.
| Free tier (F0) | 1 concurrent request — not adjustable |
|---|---|
| Standard (S0) | 100 concurrent requests by default, adjustable by support request; shared across STT and speech translation |
| Fast transcription | Requests-per-minute quota, adjustable by support request |
Published pricing
List prices as published on the vendor's own pricing page at the verification date. Follow the source link before you quote these to anyone.
| Standard real-time transcription | $1.00 per hour |
|---|---|
| Standard batch transcription | $0.18 per hour |
| Custom transcription | Real-time $1.20/hour; batch $0.225/hour; endpoint hosting $0.0538 per model per hour |
| Commitment tiers | $1,600 / 2,000 hours ($0.80/hr) through $25,000 / 50,000 hours ($0.50/hr) |
| Diarisation add-on | Real-time $0.30/hour per feature; included in batch |
| Medical tier | No medical-specific SKU published |
Data handling and compliance
The three answers a health-system security review will ask you for, as published.
| BAA | Covered by the Microsoft HIPAA Business Associate Agreement for in-scope online services |
|---|---|
| Retention | Encryption at rest documented (AES-256); retention configurable per resource |
| Training on PHI | Governed by the Microsoft DPA — get the Speech-specific position in writing |
How to integrate the Azure AI Speech API
An ordered path based on the public documentation, written for an engineering team scoping the work.
- 01
Create the resource
Provision a Speech resource in a region your customer's data-residency terms allow; region choice is hard to reverse.
- 02
Prefer Entra ID
Use token-based auth rather than long-lived subscription keys for PHI workloads.
- 03
Start with fast transcription
The synchronous transcribe endpoint is the quickest way to see clinical-vocabulary error rates on your own audio.
- 04
Customise vocabulary
Train a custom model on your specialty terminology and measure the delta; this is the main lever Azure gives you.
- 05
Watch the shared quota
The default 100 concurrent requests is shared across STT and translation — plan peak clinic load against the combined figure.
- 06
Build note structure
Azure returns transcripts. Templates, structure, hallucination controls and clinical safety are entirely your responsibility.
What the documentation does not warn you about
The judgement an endpoint list cannot give you.
Sensible when Azure is already your compliance and hosting boundary. Custom vocabulary narrows the clinical-terminology gap; it does not produce a note, and there is no medical model to fall back on.
Primary sources for this listing
Every fact above traces to one of these pages. If one has changed, tell us and the listing is re-verified.
- Speech service documentation https://learn.microsoft.com/en-us/azure/ai-services/speech-service/
- Quotas and limits https://learn.microsoft.com/en-us/azure/ai-services/speech-service/speech-services-quotas-and-limits
- Fast transcription REST reference https://learn.microsoft.com/en-us/rest/api/speechtotext/transcriptions/transcribe?view=rest-speechtotext-2025-10-15
- Pricing https://azure.microsoft.com/en-us/pricing/details/speech/
- HIPAA offering https://learn.microsoft.com/en-us/azure/compliance/offerings/offering-hipaa-us
Evidence & sources
Standards the integration steps above assume. Vendor-specific evidence is the primary source list on this page.
U.S. Department of Health & Human Services · Regulation · Tier A — primary documentation
Supports: The contractual clauses a documentation vendor's BAA must contain.
U.S. Department of Health & Human Services · Regulation · Tier A — primary documentation
Supports: Administrative, physical and technical safeguards a vendor handling recorded encounter audio must implement.
HL7 International · Standard · Tier A — primary documentation
Supports: Resource definitions (DocumentReference, Composition, Encounter, Condition, MedicationRequest) that clinical documentation output must map onto.
Source tiers are defined on the methodology page. Outbound links are unaffiliated and carry no commercial relationship.
Frequently asked questions
- Does Azure AI Speech have a public API?
- General-purpose speech-to-text with fast, batch and real-time modes and customisable vocabulary. There is no medical-specific SKU — clinical terminology is a customisation exercise.
- How do you authenticate with the Azure AI Speech API?
- Ocp-Apim-Subscription-Key header, or Entra ID bearer token
- What are the Azure AI Speech API rate limits?
- Free tier (F0): 1 concurrent request — not adjustable. Standard (S0): 100 concurrent requests by default, adjustable by support request; shared across STT and speech translation. Fast transcription: Requests-per-minute quota, adjustable by support request
- How much does the Azure AI Speech API cost?
- Standard real-time transcription: $1.00 per hour. Standard batch transcription: $0.18 per hour. Custom transcription: Real-time $1.20/hour; batch $0.225/hour; endpoint hosting $0.0538 per model per hour. Commitment tiers: $1,600 / 2,000 hours ($0.80/hr) through $25,000 / 50,000 hours ($0.50/hr). Diarisation add-on: Real-time $0.30/hour per feature; included in batch. Medical tier: No medical-specific SKU published
- Can you evaluate Azure AI Speech without a sales conversation?
- Yes — self-serve Azure resource with a free F0 tier and published prices