The real constraint: 5,000 characters per take
That’s roughly 800–1,000 words, depending on sentence length — a chapter section, not a book. A typical 80,000-word novel needs something like 80–100 takes, downloaded and joined in an audio editor.
That makes this a good fit for previewing how a chapter sounds before committing to full production, narrating shorter works (short stories, guides, single chapters), or producing a sample for a longer book. It’s a poor fit if you want one button that turns an entire manuscript into a finished audiobook — and this tool doesn’t pretend otherwise.
Choosing a narration mode
Single narrator
The default for most books — one consistent voice reading straight prose: non-fiction, guides, memoirs, anything without heavy dialogue between named characters.
Multi-speaker
For dialogue-heavy scenes, give each character a voice in the multi-speaker studio: write lines as Narrator: …, Mira: …, Tom: …. Children’s stories with a few recurring characters work especially well.
A realistic chapter workflow
- Import a section. Upload the chapter as PDF or Word, or paste the text.
- Split at natural breaks. Scene changes make better seams than mid-paragraph cuts.
- Generate each piece with the same settings. Same voice, speed and pitch keep the chapter consistent.
- Join in your editor. Use “Download all” for a ZIP, then combine the files (Audacity is free and fine for this).
More detail on stitching sections without audible seams: producing long-form audio in sections.
Publishing rights aren’t something we can decide for you
Whether AI-narrated audio can be sold, and what disclosure it needs, depends on the platform you publish through and on its current policy, which changes. Check before you commit production time.