How to enable Real-Time Avatar Streaming API in Azure Speech service?

Question

I would like to use the Azure Neural TTS Real-Time Avatar Streaming API

I already have a Speech resource created in a supported region (e.g., East US), but I understand that access to avatar streaming is currently gated and requires approval.

Could you please guide me through the steps to:

Enable the Avatar preview feature in my Azure subscription

Get access to use the real-time avatar streaming endpoint

Confirm any regions, voices, or SKUs required for this feature

I have already reviewed the official WebSocket sample and would like to test the avatar functionality using the real-time TTS + viseme stream.

Please let me know what approvals or configuration steps are required in the Azure portal to proceed.

Answer 1

Hello Ananth !

Thank you for posting on Microsoft Learn.

This functionality is in preview and not available by default. You’ll need to:

• Apply for access via Microsoft. This usually involves filling out a preview request form or contacting your Microsoft AI or Azure account representative.
• Once approved, navigate to your Speech resource in the Azure portal.
• Under Preview features (or similar), look for Text-to-Speech Avatars or Avatar real-time streaming, and switch it ON.

After approval and enabling preview:

1. Verify that you’re using the Speech SDK v1.40+ or later
2. Use regions that support avatars, and your Speech resource S0 tier
3. In code, create a WebSocket or WebRTC connection using the real-time avatar endpoint. Microsoft provides detailed samples for JavaScript, Python, C# + JS mobile clients, showing how to consume both live video and viseme (lip-sync) streams.

Links to help you :

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/text-to-speech-avatar/real-time-synthesis-avatar

https://azureaggregator.wordpress.com/2024/06/28/make-your-voice-chatbots-more-engaging-with-new-text-to-speech-features

https://www.youtube.com/watch?v=JUIs063K6z

Your resource should be located in one of these regions and uses S0 (Standard) tier.

• Southeast Asia
• North Europe
• West Europe
• Sweden Central
• South Central US
• East US 2
• West US 2

Supported voices include standard neural voices and other neural TTS voices listed in the region voice catalog).

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/text-to-speech-avatar/real-time-synthesis-avatar