How do I synthesize speech with Open AI voices?

The information in this article applies to:

• LexiconKit 10
• SpeechKit 13
• VoiceMarkupKit 10

Summary

Open AI text to speech voices are designed to generate natural sounding speech like Azure AI Speech voices. They can be easily enumerated and used with Chant tools and libraries that support Azure Speech services.

More Information

Microsoft Azure Speech services provide access to Open AI voices like Azure Speech voices. There are two model variants: Neural where voices are optimized for real-time use applications and NeuralHD where voices are optimized for quality.

Currently, Azure regions are limited to North Central US, and Sweden Central and resource must use standard pricing tier. Free tier pricing is not available for Open AI voices.

To see and hear what the Open AI voices sounds like, launch Chant Developer Workbench.

Set Azure credentials to a region and resource that is setup to access the AI Voices.

Open a Speech Synthesizer browser tab.

Select the Microsoft Cognitive Services speech API (C++ or .NET projection).

Select an OpenAI voice from the voice list.

Enter the text to synthesize and press the Start button.

Azure speech support is available for application development and Developer Workbench testing with LexiconKit, SpeechKit and, VoiceMarkupKit. Review product documentation for details:

• LexiconKit
• SpeechKit
• VoiceMarkupKit