How to Make an Audiobook From a Public-Domain Book

Public-domain books are a natural fit for AI narration. The text is free to use, the classics already have built-in search demand, and you get to build a catalog without paying for rights or waiting on an author. This guide walks through how to turn a public-domain title into a finished audiobook you can sell, and where the honest limits are.

One thing to be clear about up front: AudioProducer.ai produces the audio. You upload or paste your text, generate the narration, and download a finished MP3. We do not distribute, host, or publish to Audible, Spotify, Apple, libraries, or any podcast feed. You take the file and sell it wherever you already publish. That separation matters most with public-domain work, because the rights due-diligence is on you, not on us.

What "public domain" means (and how to verify it)

A work is in the public domain when its copyright has expired or was never granted, so anyone can use it without permission. In the United States, most works published before 1929 are public domain, and that window moves forward by one year every January. Other countries use different rules, often "life of the author plus 70 years," which means a title can be free in one country and still protected in another.

This is informational, not legal advice, and it is date-dependent and jurisdiction-specific. Before you narrate anything, verify that the exact edition you are using is actually public domain in the country where you plan to sell. Two traps catch people: a modern translation of an old work carries its own fresh copyright, and a scholarly edition with new introductions or annotations is protected even when the base text is free. When in doubt, stick to a plain, original-language text with no added editorial material, and confirm the status yourself.

Where to find clean source texts

The quality of your source file decides how much cleanup you do later. Well-known archives of public-domain literature offer plain-text and HTML versions of thousands of titles, and those are usually cleaner than a scanned PDF. Look for a text that reads as continuous prose, not one broken by page numbers, running headers, or OCR garble from a scan.

When you have a choice of formats, pick plain text or clean HTML over PDF. A good source has consistent chapter breaks, no stray footnote markers dropped mid-sentence, and no "Chapter I" repeated as an image caption. Every artifact you leave in is something the narration will try to read aloud, so a few minutes choosing the right file saves an hour of editing.

Preparing the text for narration

Classic texts need light editing before they narrate well. Strip out anything that is not part of the story: transcriber notes, license boilerplate, page markers, and any modern preface you did not intend to include (and could not include anyway if it is still under copyright). Old scans sometimes carry archaic spellings or printer conventions that sound odd read aloud, so read a few paragraphs and smooth the obvious ones.

Split the text into chapters that match the book. Clean chapter boundaries make the audio easier to navigate and easier to sell as a structured audiobook. If you want tips on cutting a long manuscript into narration-sized pieces, see how to make an audiobook chapter by chapter. Numbers, dates, and abbreviations in older texts can also trip up narration, and handling numbers and abbreviations in AI narration covers how to make those read naturally.

Casting voices for classics

A nineteenth-century novel and a children's fable want different voices. Pick a narrator voice that fits the era and tone of the book: something measured and warm for literary fiction, something brighter for a folk tale. If the book has strong dialogue between distinct characters, you can assign separate voices so conversations are easy to follow. Our guide to choosing AI voices for characters goes deeper on matching voices to a cast.

If you want to narrate a classic in your own voice, voice cloning is an option, but it requires consent: you can clone your own voice, or a voice you have explicit permission to use. You cannot clone a famous narrator or a living actor without their permission. For the mechanics of generating the actual audio, the cornerstone walkthrough on how to make an audiobook with AI covers the full flow from text to finished file.

Publishing and selling your edition

Once you download your MP3, the public-domain part gives you real freedom: you can sell your edition directly and keep the full price. Because AudioProducer.ai exports the file rather than distributing it, you publish wherever you already reach readers. Many indie creators sell public-domain audiobooks straight from their own storefront, and selling an audiobook direct on your website shows one way to do that.

Two rights questions come up often when you sell a public-domain edition. The first is whether you can copyright your specific recording, which we cover in can you copyright an AI-narrated audiobook. The second is whether it is legal to sell AI-narrated audiobooks at all, and where retailer policy differs from law, which we cover in is it legal to sell AI-narrated audiobooks. Read both before you list your edition, since public-domain source text does not answer either question by itself.

Try it on a chapter first

You do not need a card to start. The free tier gives you 1,200 words, which is enough to narrate a chapter of a classic and hear how a voice handles the era and tone before you commit. Paid plans start from 39.99 dollars per month when you want to produce a full-length title. Run one chapter, listen, adjust the voice, and then generate the rest.