PDF to speech, right in your browser.

Listening to a PDF

1. Choose your PDF above, or drop it onto the box.
2. Press play. The first time, the voice model downloads once (65 MB).
3. Click any sentence to jump there. Use the speed and voice menus in the player, or Space and the arrow keys.

What happens to the text

A PDF stores positioned pieces of text, not paragraphs. DocRead rebuilds the paragraphs from line spacing, indents and font size, then removes what you would not want read aloud: page numbers, headers and footers that repeat on every page, and hyphens that only split a word across two lines. A paragraph that runs onto the next page is joined back together.

PDFs that work best

Reports, papers, ebooks and letters exported from a word processor read well. Two-column layouts usually come out in the right order. Tables are read row by row as plain text, and footnotes are read where they appear on the page.

When a PDF won't read

• Scanned pages are images with no text layer. DocRead tells you when it finds one; run OCR on it first.
• Password-protected PDFs need the password removed: open the file in a PDF viewer and save a copy without it.
• Odd results can be fixed by hand: switch to Edit text, correct or delete what you like, then press play.

Also on DocRead: EPUB read aloud · Word to audio · Text to speech · Markdown to speech.