Turning a book into an audiobook used to mean booking a studio, hiring a narrator, and waiting weeks. With AI narration you can do it yourself in an afternoon: add your text, pick a voice, and download a finished recording. This guide walks through the whole process, what it costs, and how to handle the parts that trip people up.
What you need before you start
Just the text and an account. Your book can be a pasted draft, an exported EPUB from your writing app, or a plain .txt file. There's no software to install and no audio equipment involved. The narration runs in the cloud. If you're only testing the idea, a single chapter is enough to hear what you'll get.
Step 1: Prepare your text
Get the book into a format the tool can read: paste raw text directly, or upload EPUB, FB2, TXT, Markdown, or HTML. If you're working from a PDF, copy the text out or convert it first. PDFs are built for print, not for reading aloud, and their layout often scrambles the reading order.
Before you upload, strip anything you don't want narrated: a title page, a dedication, the table of contents, page numbers. The parser already skips most front matter, but a clean source gives a cleaner result.
Step 2: Upload or paste it
Create a new book in your library and add the text. Paste it into the text box, drop in your file, or paste a URL to import from. The parser breaks the book into chapters so each one is narrated and tracked separately. That chapter split matters later: it's the unit you regenerate if something needs fixing, so you never have to rerun a whole novel to fix one paragraph.
Step 3: Choose a narration mode and voice
This is the decision that shapes how the book sounds.
Single narrator reads the entire book in one voice. It's the default, and it's the right choice for non-fiction, articles, and most first listens. It's clear, even, and quick to produce.
Multi-character mode reads each chapter, works out who is speaking on every line, and gives each character their own voice. For dialogue-heavy fiction this is the difference between a flat read-through and something that sounds like a cast. It costs a little more because the text is analysed before it's narrated, so turn it on for the books that earn it.
Then pick your narrator voice. Every voice is available from the start. Nothing is locked behind a higher tier.
Step 4: Generate the audiobook
Start generation and let it run. Each chapter is narrated and stitched into the book in the background, so you can close the tab and come back. A short excerpt is ready in a few minutes. A full-length novel takes longer and renders chapter by chapter, so you'll often have the opening ready to listen to while the rest is still being made.
Step 5: Review and download
Listen through the result. AI narration is consistent, but no engine is perfect on every line. A name might land oddly, or a sentence might read with the wrong emphasis. When that happens, regenerate just that line or that chapter instead of the whole book. It's fast and cheap, and it's the main reason doing this yourself beats a one-shot studio recording: you can keep polishing. When you're happy, download the finished audiobook and keep the file.
How much does it cost?
You start free. Every new account gets a trial balance big enough to narrate a short chapter, so you can judge the quality on your own writing before paying anything. After that you buy credits in packs rather than a subscription, and you only spend on what you actually narrate.
Cost scales with length. A blog post is a few cents' worth of credit. A full-length novel lands in roughly the price range of a single retail audiobook. The difference is that you own the file, and regenerating it after an edit is free. Prices are shown in US dollars on the pricing page.
Single narrator or character voices: which should you pick?
Use one narrator when the book is mostly prose: non-fiction, essays, a memoir, or any draft you're proofing by ear. The even, steady read is exactly what you want for long stretches of text.
Switch to character voices when dialogue carries the story. A two-hander conversation read in a single voice is hard to follow; the same scene with a distinct voice per character is immediately clearer. The trade-off is a bit more processing time and cost per chapter, so it's worth reserving for the fiction where it pays off.
Will it sound good enough?
For most listening, yes. The narrator reads at a measured, literary pace and keeps emotion subtle, which is what holds up across hours of audio, because overacting gets tiring fast. The honest limit is high-drama fiction, where a great human performer still wins. But for self-published novels, fan fiction, non-fiction, and proofing your own drafts, the quality is well past "good enough," and the ability to regenerate weak lines closes most of the gap.
The fastest way to decide is to try it on a page of your own writing. Make your first audiobook free and listen.