Creates a prompt-generated voice for Live; no recordings or consent resource are required.
Voice creation model, or auto; omission uses the server default.
auto
Label for the saved voice.
Nonblank description of the desired voice.
Optional text spoken during creation; it must not be blank.