Multi-voice audiobook studio
Projects
Your books
A project is one book. Multi-voice cast runs each chapter through the script pass and gives every character their own voice. Single narrator voice skips all that and reads the whole book with one voice (like quick mode, but organised into chapters). Pick the mode when you create the book — it can't be changed afterwards.
Cast
Each character's voice is designed on ElevenLabs from its description (auto-filled from the persona the model inferred) the first time a chapter renders, then reused for that character across every chapter. Edit a description and re-render to change a voice. ElevenLabs refuses anything that sounds like a minor, so youthful characters are framed as “a young man” or “a cute anime female voice” — never child/kid/girl/boy.
Chapters
Bulk import
Upload a JSON file of chapters (see the local runner prompt in the README). Chapters are queued and processed in order, so the cast carries forward.