What it does
This entry points to Azure Speech in Foundry Tools (formerly Azure AI Speech), Microsoft's speech service on Azure. Its text-to-speech capability converts text into natural-sounding audio so apps and bots can speak, with realistic voices and speaking styles that can be customized to set a brand apart. The page also describes custom neural voice for building a branded voice, and prebuilt or custom avatars with natural-sounding voices. Beyond text to speech, the same service handles speech to text with captioning in more than 100 languages, speech translation, speaker recognition, and a Voice Live API for giving AI agents end-to-end speech. Developers use SDKs in languages such as C#, C++ and Java, and can run the models in the cloud or at the edge using containers. Embedded speech covers on-device use where connectivity is limited. Billing is pay-as-you-go, with text to speech charged by the number of characters converted.
Capabilities
- Text to speech with realistic, customizable voices and speaking styles
- Custom neural voice for building a branded voice
- Prebuilt and custom avatars
- Same service also covers speech to text, speech translation and speaker recognition
- SDKs, cloud service, containers and embedded on-device speech
When to use it
- Adding spoken responses to bots, voice agents or enterprise apps
- Creating a custom brand voice
- Voice features that must run on-device or in containers where cloud access is limited
- Teams already building on Azure that want speech, transcription and translation from one service
