Read PDF Aloud

Why you can’t just read a PDF aloud as it is

A PDF records where each line sits on a printed page. To a machine, the running header, the page number and the footnote at the bottom are as much text as the paragraph in the middle, so playing everything means hearing the title again every half minute.

That is why the cleaning happens before any voice starts. You hear the body text in the order an attentive reader would take it, sentence by sentence, with the furniture of the page left behind.

What gets left out, and how to bring it back

Running headers and footers, printed page numbers, footnotes, the citation markers beside words, title pages and contents are set aside. A word split by a hyphen at the end of a line is joined again, so “informa-” and “tion” are heard as one word.

A note above the text tells you how many parts were left out. Press “Show them” and they return to the reading in place; press again to hide them. Your position does not move, so checking a footnote never costs you your place.

Two columns, long reports and chapters

Journals, newsletters and conference papers often set text in two columns. Each page is taken down the left column first and then the right, which is how your eyes would take it, so no line of one column lands in the middle of the other.

Chapters come from the PDF’s bookmarks or, without them, from its headings. The chapter list shows the minutes each part runs, so a 200 page manual is as easy to dip into as a two page memo.

Your PDF read aloud in two kinds of voice

Browser voices come with your browser: free, unlimited and with no account. Voices under “On this device” start instantly and keep working offline. Online voices, such as Edge’s natural voices and Chrome’s Google voices, sound more human and need a connection, and the sentence being spoken passes through that browser maker’s service.

AI voices are the 40 narrators behind our audiobook maker, in 9 languages from English and Mandarin Chinese to Japanese and Korean; the voice menu lists those of your document’s language under “AI voices”. They need an account and run on credits: each credit covers roughly 1,000 characters, or about 500 Chinese, Japanese or Korean characters, and you are charged for what was actually read, never a rounded up passage. Replaying a passage in the same voice costs nothing for 30 days. Credit packs and plans are on pricing.

No account yet? Try an AI voice free on one passage of up to 600 characters, up to three times a day.

Highlights, speed and keyboard control

The current sentence is highlighted and, with Follow along switched on, scrolled into view. Voices that report word positions light each word as well: Windows system voices, Edge’s natural voices and every AI voice. Chrome’s Google voices highlight the whole sentence.

The speed list offers only what the chosen voice can do. Buttons move a sentence or a paragraph at a time, a slider covers the whole document, and Space starts or pauses whenever the controls have focus. Reduced motion settings are honoured, status updates are written for screen readers, and controls are 44 px on phones.

Read PDFs aloud on a phone with the screen off

For a commute or a walk, choose an AI voice. It plays as ordinary audio, so the reading carries on in your pocket with the screen locked, and play, pause and next sit on the lock screen and on your headset.

The next passages are prepared while the current one plays, so one passage flows into the next without a gap, from the first page to the last.

Where your PDF and your place are kept

The file is opened, parsed and cleaned on your device and never uploaded. It is saved in your browser’s own storage, so after a reload the page offers to pick up at the sentence you reached, in the same browser voice and at the same speed.

Before an AI voice reads a document, you tick once to confirm that you wrote it, have the right to have it read aloud, or that it is in the public domain. Its audio is stored for 30 days so replays are free; the PDF itself stays with you.

Which PDFs are ready to read aloud

A PDF exported from a word processor, a browser or a publishing tool carries its text and reads directly. A quick test: if you can select a sentence with the mouse, the reader can voice it. For a scan, a photo of each page, the page tells you at once that it found no text layer.

Want a file to keep rather than a live reading? PDF to audio turns the same PDF into an audiobook, with an MP3 for each chapter and an M4B for your player.