Convert a PDF into an audiobook you can play offline
Upload a PDF that has a text layer and its words are read aloud by an AI narrator, saved as one MP3 per chapter or as one long MP3. Scanned PDFs without selectable text cannot be read. Listen to the opening on screen for free; a PDF of up to 5,000 characters is free in full, and a longer one costs $5, paid once, for up to 600,000 characters.
- Works only on PDFs where you can select and copy the text
- New chapters start at lines like "Chapter 7", "CHAPTER VII", "Part Two" or "Prologue"
- Words split across two lines with a hyphen are joined back together
- Page numbers, running headers and footnotes may be read aloud
- One PDF up to 50 MB; an EPUB or Word copy of the same text usually sounds cleaner
Check the text layer first
Open the PDF and try to highlight a sentence with your cursor, or press Ctrl+F (Cmd+F on a Mac) and search for a word you can see on the page. If the words highlight or the search finds them, the PDF has a text layer and can be read. If you can only draw a box over the page as if it were a photo, it is a scan, and the upload will tell you it cannot be read.
A scan needs text recognition first. Many PDF editors have an OCR or "make searchable" option; once the text can be selected, upload the new file.
What gets read from a PDF
- Everything in the text layer, page by page, including page numbers, running heads with the book or chapter title, and footnotes at the bottom of a page
- Single-column pages come out in the right order; two-column layouts, common in papers, can mix lines from both columns
- Tables and figure captions are read as plain lines, which rarely sounds natural
- Pictures and charts are skipped, because there are no words in them to read
- If no chapter lines are found, the whole PDF becomes one long track
When another file is a better choice
A PDF is a printed page frozen in place, and its text carries the print layout with it. If the same book or report exists as an EPUB or Word file, upload that instead: the chapters come from real headings and no page number lands in the middle of a sentence. For your own documents, saving as DOCX rather than PDF gives the cleanest audio.
For a report or paper with parts you do not need, such as references or appendices, cut them from a copy of the PDF before uploading. The character limit counts everything in the file, so a long bibliography uses up allowance too.
How the tool works, what it costs and how your files are handled is on the Text to Audiobook page; prices for every tool are on the pricing page.
Last updated
Questions and answers
Can ChatGPT read a PDF out loud?
It can take a PDF you attach and summarize it, and voice mode can read short replies aloud. It does not turn a whole PDF into audio files, and with a long book it tends to summarize instead of reading every page.
How do I know whether my PDF will work?
Try to select a line of text or search for a word in it. If that works, the PDF has a text layer and can be converted.
Can it read a scanned PDF?
No. A scan is a picture of a page with no text inside. Run OCR on it first, then upload the searchable version.
Will page numbers be read out?
They may be, along with running headers and footnotes, because they are part of the PDF text. An EPUB or DOCX of the same text avoids that.
How much does it cost to convert a PDF?
A PDF of up to 5,000 characters is free in full. A longer one is $5, paid once, for up to 600,000 characters, about 100,000 words. Unlimited costs $19 a week and allows 1,500,000 characters per job.
Is my PDF kept private?
Only you can open the job. The text is sent to an AI model provider only to be read aloud and is not used to train models. Jobs started without an account are deleted after 7 days, and you can delete one yourself at any time.