Microsoft Azure → Amazon Web Services
AWS equivalent of Azure Speech
The closest AWS equivalents of Azure Speech are Amazon Transcribe, Amazon Polly, Amazon Transcribe Call Analytics, Amazon Nova and AWS HealthScribe.
About Azure Speech
| Provider | Microsoft Azure |
|---|---|
| Status | Generally available Custom neural voice and some avatar features are Limited Access. |
| What it is | Speech to text (real-time, fast, and batch transcription, custom speech), neural text to speech (custom voice), speech translation, pronunciation assessment, avatars, Voice Live, and edge containers. |
| Documentation | learn.microsoft.com |
| Pricing | Official pricing page |
AWS mappings (5)
| Capability | AWS equivalent | Match |
|---|---|---|
| Speech to text | Amazon Transcribe | Direct match |
| Text to speech | Amazon Polly | Direct match |
| Contact-center call analytics | Amazon Transcribe Call Analytics | Partial match |
| Real-time speech-to-speech AI | Amazon Nova | Partial match |
| Ambient clinical documentation (AI scribe) | AWS HealthScribe | Partial match |
Speech & conversational AI(4)
Speech to text Direct match
The Azure equivalent of Amazon Transcribe is Azure Speech speech-to-text.
- Azure offers real-time, fast (synchronous file), and batch transcription, custom speech models, pronunciation assessment, and edge containers.
- Transcribe offers streaming and batch, custom vocabularies and custom language models, speaker partitioning, PII redaction, and toxicity/content filtering. It's HIPAA-eligible.
- Foundry also offers OpenAI speech-to-text models (for example gpt-4o-transcribe family) as an LLM-based alternative.
Text to speech Direct match
The Azure equivalent of Amazon Polly is Azure Speech text-to-speech.
- Polly voice engines are generative, long-form, neural, and standard. Polly supports SSML and lexicons.
- Azure has neural voices, batch synthesis, and avatars, plus custom neural voice (brand voice), which is Limited Access and needs Microsoft approval.
Contact-center call analytics Partial match
Amazon Transcribe Call Analytics maps partially to Azure Speech batch transcription combined with Azure Language or Foundry models. Microsoft documents this as a call-center scenario rather than a single API.
- Call Analytics is one API with sentiment, issue detection, rule-based categories, and generative call summaries.
- On Azure you compose Speech (transcribe, diarize, redact) with Language or Foundry models for sentiment and summary, or use Content Understanding's call-center analyzers.
- For a full contact center, Microsoft's comparison maps Amazon Connect to Dynamics 365 Contact Center.
Real-time speech-to-speech AI Partial match
Amazon Nova Sonic (speech-to-speech on Bedrock) maps to Azure Voice Live and voice-based prompt agents in Foundry (preview), which also support realtime audio models.
- Nova 2 Sonic is a single speech-to-speech foundation model you call through Bedrock's bidirectional streaming.
- Voice Live is a managed orchestration service (speech recognition, turn-taking, model interaction, synthesis) that works with realtime models or a cascaded STT → LLM → TTS chain. Foundry voice agents add telephony and avatar options (preview).
Industry & specialized ML(1)
Ambient clinical documentation (AI scribe) Partial match
Azure has no documented one-API equivalent of AWS HealthScribe. You can compose it from Azure Speech, Text Analytics for health, and Foundry models (including Microsoft healthcare AI models).
- HealthScribe is one HIPAA-eligible API that returns a transcript, a structured clinical note, and evidence links. AWS also now offers the Amazon Connect Health ambient agent.
- Microsoft's commercial ambient-scribe product (Dragon Copilot) is an application, not an Azure API, and wasn't evaluated in this release.