Bring your own voice

We are excited to be introducing voice cloning on Wit.ai, where you can use your own custom voice with our existing text-to-speech service.

We are starting to open access to this feature for Voice SDK developers who are interested in voice cloning on Wit.ai. This will allow developers to create their own voice for use in our existing POST /synthesize API by uploading 15–30 minutes of voice samples as a reference. This means you can craft a specific voice that will work for nearly any scenario, and ensures it will be unique to your app.

Once you have been granted access to the feature, you will be able to upload your voice data via the voices page on Wit.ai. After the data is uploaded and validated, your voice will be ready to use as soon as the next day.

For Voice SDK developers who are interested in using this feature, go to the voices page on Wit.ai to request access.

We look forward to seeing developers bring their own voices to use in our text-to-speech service and seeing the endless customization possibilities that are created with this new feature. Feel free to contact us with any questions or comments.

Wit.ai Team