How to Turn a Reedsy Book Into an Audiobook With AI

Reedsy Book Editor is where a lot of indie authors write and lay out their books, because it is free, runs in the browser, and turns a manuscript into clean EPUB and print-ready files. What it does not do is make an audiobook. If you want your Reedsy book read aloud, you need a path from the editor to a finished audio file. This post walks that path from start to finish, so you can take a book you already wrote in Reedsy and come out the other side with an MP3 you can publish wherever you already sell.

The short version: export your text out of Reedsy, bring it into AudioProducer, keep your chapter breaks intact, choose how each part should sound, and generate the audio. You download the file and take it from there.

Exporting from the Reedsy Book Editor

Reedsy exports a polished EPUB and a print-ready PDF from the same project. For audio you want the EPUB, not the PDF. The PDF is laid out for the printed page, with fixed line breaks, running heads, and page numbers baked in, and all of that gets in the way when a tool tries to read the text in order. The EPUB carries your words and your chapter structure without the print-only scaffolding.

To export, open your book in the editor, go to the export options, and choose the EPUB format. Reedsy generates the file and emails you a download link or hands it to you directly, depending on how your project is set up. Save the EPUB somewhere you can find it. That single file is everything the next step needs.

If your book is still in progress and you only want to hear one chapter, you can export the whole project anyway and just work with the part you care about once it is imported. There is no need to split the file yourself before exporting.

Cleaning up the text for the ear

Text written for the eye and text meant for the ear are not always the same thing, and a quick pass before you generate audio saves you from re-recording later. Front matter is the first thing to decide on. A title page and a dedication are fine to keep if you want them narrated, but many authors trim the copyright notice out of the audio version because a block of legal boilerplate is a slow way to open a listening experience.

Then look for anything that only works on the page. Footnotes, endnotes, and cross-references like "see the table on page 40" do not translate to audio, so fold the important ones into the body text and drop the rest. The same goes for URLs spelled out in full and for scene-break glyphs, which a narrator would either skip or read as a literal symbol. The goal is a document that reads top to bottom the way you want it heard.

Chapter-by-chapter workflow

One advantage of starting from a Reedsy export is that the chapter structure is already clean. When you import the EPUB, those chapters should come across as distinct sections rather than one long wall of text. Keeping chapters separate matters for two practical reasons. It gives listeners natural stopping points, and it lets you regenerate a single chapter later without rebuilding the entire book.

After importing, check that each chapter heading arrived as a heading and not as an ordinary line of body text. If a break got flattened somewhere in the export, fix it in the editor before you generate any audio. Getting the structure right once means every later step, from casting voices to the final download, lines up with the book you actually wrote. It also makes the chapter-at-a-time approach easy: run one chapter through first, listen to it, and only then commit the rest.

Voice selection and character casting

With the text in and the chapters clean, you decide how the book sounds. You can narrate the whole thing in a single voice, which is the standard for most fiction and nearly all nonfiction. If your book leans on dialogue and you want each speaker to come through distinctly, you can assign different voices to different characters. Casting a voice per character and rendering in one pass sits between a plain solo reading and a full studio production, and it works well for dialogue-heavy stories.

If you would rather narrate in your own voice, voice cloning is available with one firm rule: you can only clone a voice you have the right to use. That means your own voice, or a voice whose owner has given you permission. We work on AudioProducer, so we are not a neutral party here, but the consent rule is not something we bend.

What you export and where it goes

When the narration sounds the way you want, you export the finished audio as an MP3 and download it. That file is yours. AudioProducer produces the audio; it does not publish or distribute it for you, and it does not push it to any store or feed. You take the MP3 and upload it to whatever platform you already use, the same way you would take the EPUB and print files Reedsy gave you and send them where they need to go.

If you are new to the whole process and want the wider picture rather than just the Reedsy-specific steps, our guide on how to make an audiobook with AI covers the fundamentals. And if your next project lives in a different tool, the import path is the same for a Google Doc, a Word manuscript, a Vellum book, or a Scrivener project, so you are not tied to one writing app.

You can try the full flow on the free tier, which gives you 1,200 words at no cost and no card, enough to run a chapter through and hear how your book sounds before you decide. Paid plans start from $39.99 per month when you are ready to take a whole book through.

FAQ