MotionShineShort answer: open the PDF in the free PDF to audio tool. It pulls out the text in reading order, splits it into parts at the headings, leaves out page numbers and running headers, and reads each part with an AI voice into a tagged MP3. Save the parts as a ZIP or as one MP3 with chapter marks, and copy them to your phone. It runs in your browser, so the PDF is never uploaded. A scanned PDF needs OCR first, and for an EPUB or Word file use the audiobook maker.
Both tools are ours, so weigh what we say about them as a vendor's own words. Facts about other apps are from their makers' pages, checked on 7 November 2026.
Step 1: check that the PDF has real text
This decides everything else. A PDF made from Word, a website or an ebook has text inside it. A PDF made by a scanner or a phone camera is a picture of pages, and no voice can read a picture.
The quick test: open the PDF in Preview on a Mac (or any PDF reader) and try to select one sentence with the mouse. If the words highlight, it has text. If you can only draw a box over the page, it is a scan.
For a scan, get the words out first with PDF to text, which has OCR, then paste the text into the audiobook maker. Check names and numbers afterwards, because OCR misreads them most.
Step 2: make the audiobook with PDF to audio
- Drop the PDF on the PDF to audio page. It shows the page count and a list of parts.
- Look through the list. Each row shows the word count and how long it will take to hear. Untick what you do not want read, such as the contents page or the references, and type over any title to rename it.
- Fill in Title and Author. They go into the MP3 tags.
- Pick a Voice in the PDF's language and press Hear to listen to the first lines. There are 20 voices in nine languages: English (American and British), Chinese, Spanish, French, German, Russian, Arabic, Hindi and Bangla.
- Set the Speed (0.5x to 2x) and the Pauses after sentences and between paragraphs.
- Press Read it aloud. Each part gets a player and an MP3 link when it is ready, and you can listen to part 1 while the rest is made.
- Press Download all (ZIP) for every part plus a playlist, or One MP3 with marks for a single file with a chapter mark at each part.
The first time, the page downloads the voice, about 92 MB for English, and keeps it for next time.
How the PDF becomes chapters
The tool reads text size to find headings. The two largest heading sizes start a new part, and smaller headings stay inside the part around them. If a part runs past about 3,750 words, or the PDF has no headings, it is cut into parts of about 2,500 words, always at the end of a paragraph.
On the way it cleans up the things that make PDFs sound bad read aloud:
- Lines that repeat at the top or bottom of most pages, such as running headers, footers and page numbers, are left out.
- A word broken with a hyphen at the end of a line is joined back together.
- Pictures are skipped. Tables are read cell by cell, and maths may come out as letters and symbols.
Each MP3 is named with a number and the part title, such as 01 Introduction.mp3, and tagged with the title, author and part number, so music and audiobook apps play them in order. The ZIP also holds a playlist.m3u file.
How long it will be, and how big
The tool counts about 150 spoken words a minute at 1x speed. Its own examples:
| Text | Words | At 1x | At 1.2x or 1.3x |
|---|---|---|---|
| A report | 6,000 | about 40 minutes | about 31 minutes at 1.3x |
| A short story | 7,500 | 50 minutes | about 42 minutes at 1.2x |
| A novel | 90,000 | 10 hours | about 8 hours 20 minutes at 1.2x |
The MP3s are saved at 64 kbps unless you pick 128 kbps. 64 kbps is 480 KB a minute, so the 40 minute report comes to about 19 MB. Making the audio takes longer than listening on many computers, because your own computer does the work. The tool's page says the English voices made about one second of speech every two seconds on a Mac, so keep the tab open and the Mac awake for a long book. If you stop, finished parts are kept in the browser and come back when you open the same PDF.
Other free ways, and what each gives you
| Way | Makes a file | Chapters | Uploads your PDF | Cost |
|---|---|---|---|---|
| PDF to audio | MP3 per part, or one MP3 with marks | Yes, from headings | No | Free |
| Preview: Edit > Speech > Start Speaking | No, listening only | No | No | Free on a Mac |
say -f book.txt -o book.aiff in Terminal | AIFF, one file | No | No | Free on a Mac, needs the text as a .txt first |
| NaturalReader free plan | No: its help center says the free plan "Cannot convert to mp3" | No file | Yes, cloud | Free plan, paid for MP3 |
| ElevenReader free plan | No: "audio files can't be exported from the app" | In the app | Yes, cloud | 10 hours of listening a month free |
Preview and the say command are covered step by step in how to use text to speech on Mac. For a wider list of voice tools and their free limits, see the best free text to speech software.
The honest limits of a free voice
- It is a reading, not a performance. One voice reads the whole book at an even pace. It does not act out characters or change tone for a joke. For a voice per speaker in a script, use multi-voice text to speech.
- Names get guessed. Words that are not in the voice's dictionary are guessed. Add them under Pronunciation fixes with the way they sound, such as Kee va for Caoimhe, and press the speaker button to check.
- Mostly-sentences PDFs work best. Slides, forms, footnote-heavy papers and pages of tables sound choppy. Untick those parts.
- The speed is baked in. Set it before you make the audio, or use the playback speed in your podcast or audiobook app later.
Whose books can you turn into audio?
Your own writing, public domain books and texts you have the rights to are fine to share. The tools only read files without DRM, and an audiobook of someone else's book is for your own listening. If you want to sell an AI-narrated audiobook of your own work, the tool pages note that each voice's licence allows commercial use (some ask for credit), and that each store has its own rules on AI narration, so read the store's rules before you publish.
Make it a video instead
If the PDF is a guide or a slide deck you want to walk people through, a screen recording with your own voice may serve better than an audiobook. MotionShine, by the maker of this site, is a free Mac screen recorder with auto zoom and a full editor. Free exports go up to 720p at 30 fps with no watermark. How to record your voice on a Mac covers the sound side.
FAQ
Can I convert a PDF to an audiobook for free?
Yes. The PDF to audio tool on this site is free with no account and no page limit, and saves MP3 files. On a Mac, Preview can read a PDF aloud for free but does not save a file.
How do I turn a scanned PDF into audio?
Run it through OCR first, because a scan is a picture with no text in it. Use PDF to text, which has OCR, then paste the text into the audiobook maker and make the MP3s.
Can I listen to a PDF on my phone?
Yes. Save the MP3 parts or the ZIP and copy them to your phone. MP3 plays in almost every music, podcast and file app, and the tags keep the parts in order.
Will it read page numbers, headers and footnotes?
Page numbers, headers and footers that repeat on most pages are left out. Footnotes and reference lists are read if they are in the text, so untick those parts before you start.
How long does it take to turn a PDF into audio?
It depends on your computer, and it is often slower than listening. You can play the first part as soon as it is ready while the rest is made.
Can I get one audiobook file instead of many?
Yes. One MP3 with marks joins the finished parts into one MP3 with a chapter mark at each part. Apps that read MP3 chapter marks let you jump between them; others play it from start to end.
