Text-to-Speech currently supports both standard and neural voices. However, since the neural voices provide more natural sounding speech output, and thus, a better end-user experience, we are retiring the standard voices on 31st August 2024 and they will no longer be supported after that date.
Here is a quick comparison between Standard Text-to-Speech and Neural Text-to-Speech features:
Standard text-to-speech
Neural text-to-speech
- Noticeably robotic
- Natural sounding, closer to human-parity
- Limited capabilities in voice tuning
- Advanced capabilities in voice tuning
- No new investment in future voice fonts
- On-going investment in future voice fonts
Timeline
Phase
Date
Description
Announcement
Aug 17, 2021
Deprecation was announced
Deprecation
Aug 31, 2024
Standard voices will no longer be supported
Impact
Azure Cognitive Services' Standard Text-to-Speech is retiring on 31 August 2024, migrate to neural voices.
Required Action
A migration guide is available in the documentation.
Here's the official report from Microsoft:
Review the price structure and listen to the neural voice samples to determine the right voice for your business needs.
To make the change, follow the steps to update the voice name in your speech synthesis request to the supported neural voice names in chosen languages by 31 August 2024. Starting 1st September 2024, standard voices will no longer be supported, please use neural voices for your speech synthesis request, on cloud or on prem.
Contact
You can get in touch through the following options:
Azure Cognitive Services' Standard Text-to-Speech is retiring on 31 August 2024
Deadline: Aug 31, 2024 Impacted Services:
More information:
Notice
Here's the official report from Microsoft:
Timeline
Impact
Azure Cognitive Services' Standard Text-to-Speech is retiring on 31 August 2024, migrate to neural voices.
Required Action
A migration guide is available in the documentation.
Here's the official report from Microsoft:
Contact
You can get in touch through the following options: