PDF to Audio
PDF to Audio reads a PDF aloud with 28 Kokoro-82M English voices in your browser, leaving out running headers and page numbers, and exports the reading as an MP3 or WAV audiobook with no page limit and no upload.
At a glance
- Accepts
- A PDF with a text layer (scans need OCR PDF first)
- Outputs
- Editable cleaned text, live playback, and a 24 kHz mono MP3 (64 or 128 kbps) or 16-bit WAV
- Where it runs
- Entirely in your browser, on your device
- What leaves your device
- Nothing - the PDF and its text never leave your device; the voice model is downloaded to you
- Limits
- English only; scanned PDFs have no text to read; tables, footnotes and columns are read in the order the PDF stores them; a whole book takes a long time on a phone, and the audio is held in memory at about 2.9 MB per minute
- Account
- Not required - there is no sign-up
- Price
- Free, with no watermark and no usage cap
- Made by
- Visrut
- Last updated
- 20 October 2026
Drop in a PDF and listen to it: a book, an article, a report, your lecture notes. The text is pulled out of the PDF, tidied up for listening, and read aloud by one of 28 natural English voices. Playback starts after the first sentence, and when the reading is done you can download it as an MP3 audiobook. It is free, there is no page limit, and the PDF never leaves your device.
How to use it
- Drop a PDF on the page, or click to choose one.
- Pick the pages to read, if not all of them, and check the text. Cut anything you don't want to hear.
- Choose a voice and speed, then press Read aloud.
- Listen as it reads, or download the MP3 when it is done.
Tidied up for listening
A PDF stores text as positioned pieces, not as prose, so reading it out as-is would have the voice say every running header and page number, and stumble over words split across lines. Before anything is read, lines that repeat at the top or bottom of most pages are left out, bare page numbers go, hyphenated words are joined back together, and lines become sentences - including sentences that run on to the next page. The result is in an editable box, so you have the last word.
Honest limits
- Scanned PDFs need OCR first. A scan is pictures of pages; run it through OCR PDF to get text.
- English only. The in-browser voice model only includes English pronunciation.
- Tables, footnotes and columns are read in the order the PDF stores them, which is not always the order you would read them. Cut or move them in the text box.
- A whole book takes a while. Making the audio runs on your device, so keep the tab open; the page keeps the screen on while it reads and asks before you leave.
Frequently asked questions
Is my PDF uploaded anywhere?
No. The PDF is read with pdf.js and spoken by the Kokoro voice model, both running in your browser tab. The only downloads are the voice model files, the first time you use them; your PDF and its text are never sent anywhere.
Can it turn a whole book into an audiobook?
Yes, there is no page or length limit. A 300-page novel is roughly 9 to 10 hours of audio. How long it takes to make depends on your device: a laptop with a working graphics chip (WebGPU) is usually many times faster than real time, a phone's processor can be slower than real time. The page measures your device's speed after the first sentences and shows the time left. For a long book, a chapter at a time is kinder to a phone, and the finished audio is held in memory (about 2.9 MB per minute) until you download it.
Does it read the page numbers and headers out loud?
It tries not to. Lines that repeat at the top or bottom of most pages (a book's running title, a report's footer) and bare page numbers there are left out, words hyphenated across a line break are joined back together, and lines are joined into sentences. The text appears in an editable box before anything is read, so you can cut anything else, like footnotes or a table of contents.
My PDF is a scan. Why is there no text?
A scanned PDF is pictures of pages, with no text inside it to read. Run it through OCR PDF first, which recognises the text in your browser, then bring the new PDF here.
Which voices and languages are there?
28 English voices, American and British, female and male. The browser version of the Kokoro model only includes English pronunciation, so a PDF in another language will be read with English pronunciation, which is not useful.
What do I get?
An MP3 (64 or 128 kbps) or a lossless WAV of the whole reading, made in your browser. You can also just listen: playback starts as soon as the first sentence is ready.
Does it work with password-protected PDFs?
Not directly. Unlock it first with Remove PDF Password (you need to know the password), then bring the unlocked file here.