ToolsCompaniesAbout

Azure AI Speech

by Microsoft@microsoftRedmond, Washington, USANo ratings yet

Speech-to-text, text-to-speech and speech translation APIs in Microsoft Foundry Tools

Visit azure.microsoft.com
Tool type
Speech AI APIs and SDKs
Made for
Voice application developers
Runs on
Azure APIs, containers and devices
Callable via REST APIs and SDKs; deployable as containers in the cloud or at the edge, and embedded speech for on-device scenarios.
Last checked
28 Oct 2026
2 product sources

What it does

Azure Speech takes in audio, text, or voice input and returns transcripts, synthesized speech, translated speech or text, and speaker recognition results. Developers call its APIs or use SDKs in languages such as C#, C++ and Java to add speech capabilities to applications. The service also supports custom speech models adapted to specific vocabulary or acoustic conditions, custom neural voices, and avatars.

People typically use Azure Speech to build voice-enabled agents, transcribe call center or meeting recordings, add spoken responses to bots, and translate live audio into other languages. Distinct options include embedded speech for on-device use where connectivity is limited, container deployment to run models in the cloud or at the edge, and the Voice Live API for real-time voice agents, which the product page describes as in public preview.

Azure Speech suits developers and enterprises building speech features into apps or contact-center workflows. Use requires an Azure account; a free tier provides limited audio hours and characters per month, while most usage is pay-as-you-go by audio hour or character. Limitations include the need for Azure setup, quotas that vary by tier, and custom voice and personal voice being restricted or limited-access features.

What it can do

  • Speech-to-text transcription, including real-time and batch modes
  • Text-to-speech synthesis with prebuilt and custom neural voices
  • Speech-to-speech and speech-to-text translation of audio streams
  • Speaker recognition and pronunciation assessment
  • Custom speech models adapted to vocabulary and acoustic conditions
  • Container deployment for cloud or edge, and embedded speech for on-device use

Works with

  • Azure OpenAI in Foundry Models (for GPT-based solutions and Whisper transcription)
  • Azure Content Understanding (for post-call analytics)

Example workflows

Call center transcription and analytics
Input: recorded customer calls. Steps: send audio to speech-to-text transcription, then pass transcripts to post-call analytics in Azure Content Understanding. Output: text transcripts and analytics insights about the calls.
Voice-enabled bot reply
Input: user text from a chat app. Steps: send text to text-to-speech with a chosen prebuilt or custom neural voice and play the audio. Output: spoken audio response in the bot's voice.

Before you choose

  • Pricing is metered per audio hour for speech-to-text and translation, per character for text-to-speech, and per transaction for speaker recognition; costs grow with volume.
  • Custom voice and personal voice are limited-access features; personal voice requires application approval for pre-approved use cases.
  • Voice Live API and voice agents are described as in public preview, so features may change.

Listed under

Get started

Create an Azure account, then use Azure Speech resources and the documented APIs and SDKs.

  1. Create or sign in to an Azure account at the Azure portal.
  2. Create an Azure Speech resource in Foundry Tools and note its endpoint and key.
  3. Install an Azure Speech SDK for your language (for example C#, C++ or Java) and call the speech-to-text or text-to-speech APIs.
Get started with Azure

Plans and prices

Prices in USD

Speech to text and translation are billed per second of audio; text to speech per character; custom endpoint hosting per model per hour. Unit prices are masked in the reviewed information, so check the pricing page or calculator for current rates.

Free (F0)FreeFree monthly allowance per the F0 table; not a permanent unlimited free tier.
  • 5 audio hours of real-time transcription per month, shared between Standard and Custom
  • 0.5 million characters of Neural text-to-speech per month
Pay as You GoUsage-basedNamed meter; unit prices not readable in the reviewed information.
  • Metered speech-to-text, translation and text-to-speech usage

Source: azure.microsoft.com/en-us/pricing/details/cognitive-services/speech-services, checked 28 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 28 Oct 2026. Each source supports part of the overview; availability and prices can change.

Ratings and reviews

No ratings for Azure AI Speech yet.

Be the first to rate it

What’s changed

Free account

See what’s changed at Azure AI Speech

Launches, price moves and pivots, each with a date and a source. Free with an account.

Signals

Free account

See Azure AI Speech’s signals

Launch traction, hiring and release activity, tracked over time. Free with an account.

Data and security

Free account

See how Azure AI Speech handles your data

Training on your files, compliance, hosting and the models behind it. Free with an account.

Compare with similar tools

Free account

Compare Azure AI Speech side by side

Prices, platforms and features against similar tools, with sources. Free with an account.

Where this comes from

How we research
Company website
azure.microsoft.com: product, pricing and policy pages
YC directory mirror
Batch and company name
Web search
Confirms the official site
Social accounts
Accounts the company links, and its posts on X
Human review
Checked by a person
knowledge_base

Last checked 28 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.