hey guys, is there a way to generate a preview of the assistant voice from a script via the api? For eleven labs I'd hit their endpoint but what about vapi voices?