Voice cloning workflow

Create the voice in the dashboard

Start on the Voices page, choose to create a voice and upload a clear recording of one speaker. Use your own recording or obtain the speaker’s explicit permission to clone their voice. Complete the permission confirmation and voice details, then wait for encoding to finish.

Creation and encoding are automatic; these steps do not include human review. Once ready, copy the voice ID and use it as the voice parameter in speech requests. The voice belongs to your account; your application does not need to upload the original recording with each script.

This page documents the dashboard workflow and use of the resulting voice. It does not define an OpenAI-compatible public voice-creation endpoint.