Articles

How to make your Mac read a PDF aloud

Updated August 2026 · 6 min read

Your Mac has been able to read text aloud since before most of us owned one. The tools are just buried. Here is every built-in way to make macOS read a PDF to you, how to make the voice bearable, and what to reach for when the built-ins run out of road.

Speak Selection: the two-key method

The fastest built-in route is Spoken Content's Speak selection:

  1. Open System Settings → Accessibility → Spoken Content.
  2. Turn on Speak selection. The default shortcut is Option + Esc; you can change it here.
  3. Open the PDF in Preview (or any app), select some text, and press the shortcut.

macOS speaks the selection and shows a small floating controller with pause, skip and speed buttons. Press the shortcut again to stop. If you press it with nothing selected, some apps will speak the whole visible document, but results vary by app.

Make the voice worth listening to

The default system voice is fine for alerts and rough for chapters. In the same Spoken Content pane:

  • Click System voice and browse the list. The Siri voices are the best of the built-ins; macOS downloads them on first use.
  • Drag Speaking rate up a notch. Most people settle around 1.1 to 1.3 times normal speed once their ears adjust.

Even the best system voices, though, read sentence by sentence with flat intonation. They do not carry the rhythm of a paragraph the way modern neural narrators do, which matters more the longer you listen.

The terminal party trick: say

Every Mac also has a command-line voice. Open Terminal and try say "hello there". Pointed at a text file, say -f notes.txt -o narration.aiff will render the whole thing to an audio file. It cannot read a PDF directly - you would need to extract the text first - and the output uses the same system voices, so treat it as a fun tool for scripts rather than a serious audiobook pipeline.

Where the built-ins fall short

All of these share the same limits, because none of them actually understands documents:

  • No follow-along. The page does not highlight what is being read, so switching between listening and reading means hunting for your place.
  • Everything gets read. Headers, footers, page numbers and citation brackets are narrated as if they were sentences.
  • Column chaos. Two-column layouts often read straight across the page, splicing unrelated sentences together.
  • Scans are silent. A scanned PDF has no text layer, so there is nothing to speak - you need OCR first (here is how).
  • No memory. Stop for the day and tomorrow you start by finding yesterday's sentence yourself.

When you want a real narrator

If you listen to more than the occasional page, a dedicated reader changes the experience: neural voices that sound like a person, paragraph highlighting locked to the audio, automatic skipping of page furniture, OCR for scans, per-book resume, and playback speed that remembers itself. That is what Unknown does, entirely on-device - your PDFs never leave your Mac. For the full walkthrough, see how to turn a PDF into an audiobook on Mac.

Try it

Unknown reads your PDFs and EPUBs aloud in 28 lifelike voices and lights up every paragraph in time with the audio. Free for 50 pages a month.

Download Unknown for Mac →

100% on-device · macOS 14.1 or later

FAQ

Why is my Mac not reading my PDF aloud?

Check three things: Speak selection is enabled in System Settings, you actually have text selected when you press the shortcut, and the PDF has a text layer. If you cannot select text in the document, it is a scan and needs OCR first.

Can macOS read a PDF while I do something else?

Yes. Once Speak selection starts, the narration keeps going while you switch apps. It stops at the end of the selection, though, so you will be back to reselect before long.

What is the keyboard shortcut to make a Mac talk?

Option + Esc speaks the current selection, once Speak selection is turned on in System Settings → Accessibility → Spoken Content.

Does any of this need the internet?

No. Spoken Content, say, and Unknown all generate speech locally on your Mac. Siri voices need one download, then work offline.