How Do You Get a Mac to Read Text Aloud (and Keep Reading)?

To make your Mac read text aloud, open System Settings, click Accessibility, then Spoken Content (called Read & Speak on macOS Tahoe 26 and later) and switch on Speak selection. Select some text and press Option-Esc. Press it again to stop. It's free and built in, and it's shaped for passages, not long articles.

You've got a 5,000-word article open in a tab. You've read the first two paragraphs twice, because the words slid off and you had to go back. Your eyes are doing the thing where they move and nothing lands.

So you type "text to speech mac" into a search bar, which is how you ended up here.

Maybe you've tried the Mac's voice before. You selected a paragraph, pressed something, heard a voice that sounded like a satnav from 2009, and closed it. Or it worked, sort of, and then it read the cookie banner, the navigation menu and the "related articles" before it reached the article. Or you stopped to answer a message, came back, and had no idea where the voice had got to. You scrolled around trying to find the sentence you'd last heard, and gave up, and the tab went back to being a tab you'd read later.

It's 4pm. The article is one you meant to read on Tuesday. Someone sent it with "thought of you", and every time the tab catches your eye there's a small drop in your stomach, because it's one more thing on a list you're already failing to clear.

Listening feels like the way out. You'd walk round the flat with it, or sit with your eyes shut. So you try. And within five minutes you're doing the same old thing, except now there's a robot voice involved, and you're more annoyed than when you started.

That's the moment that matters. Not the first press of the shortcut, but the third or fourth time, when you decide whether this thing is for you.

Most people decide it isn't. They conclude Mac text-to-speech is bad, or that they're not the sort of person who can listen to articles, or that the problem is their focus. Some of that is a reasonable verdict on a built-in tool that was never built for the job you've given it. None of it is a verdict on you.

Why does this feel so close to working, and then not?

Because the Mac's built-in reading tools are good at one shape of task and not the other.

The shape they're good at: tell me what this says. A paragraph. An email. A dialog box. A sentence you want to hear in a different voice to catch a typo. You make one request, the Mac gives you one answer, you move on.

The shape they struggle with: read with me for forty minutes. That's a session. A session needs a place you can come back to, a voice you can tolerate across thousands of words, a speed you can change without breaking your concentration, and some way for what you heard to still be in your head tomorrow. Call it the session gap. It's the distance between a tool that answers a request and a tool that holds a sitting.

I'll come back to the gap. First, the practical part, because the built-in tools are better than most people give them credit for, and you may find the answer is a two-minute fix.

How do you turn on Speak selection (and what's it called on your Mac)?

Speak selection is the one most people want. Here's the path, verified against Apple's Mac User Guide page "Have your Mac speak text that's on the screen":

1. Choose Apple menu, then System Settings.
2. Click Accessibility in the sidebar (you may need to scroll down).
3. Click Spoken Content on macOS Sequoia 15, or Read & Speak on macOS Tahoe 26 and later.
4. Turn on Speak selection.
5. Select text anywhere and press Option-Esc. Press it again to stop.

Apple appears to have renamed the pane. Apple's pages for Sequoia 15 call it Spoken Content. Its pages for Tahoe 26 and later call it Read & Speak. Everything inside works the same way. If you're on an older Mac that still has System Preferences instead of System Settings, the idea is the same but the menu labels differ, and I haven't checked those versions.

Two details from the same Apple page save a lot of fiddling.

With nothing selected, it reads the window. Apple says that if no text is selected, "available text items in the current window are spoken." So you don't have to drag-select a whole article. Click into the page and press the shortcut.

The shortcut can change. Click the Info button beside Speak selection to pick a different key combination. That same panel lets you choose whether words, sentences or both are highlighted as the Mac speaks, and set the highlight colours. If you've been listening with no idea where the voice is, turning highlighting on is the first fix.

Apple also says "the controller is especially useful when your Mac is reading long pieces of text." The buttons depend on your version. On Tahoe 26, Apple lists six: speaking rate, skip back, play or pause, stop, skip forward, and a button that opens Accessibility Reader. On Sequoia 15, Apple lists decrease rate, skip back a sentence, play or pause, stop, skip forward a sentence, and increase rate, with no Accessibility Reader button. You choose when the controller appears from the same Info panel.

In many apps there's a second route: Edit, Speech, Start Speaking, then Edit, Speech, Stop Speaking. Apple lists it on the same page.

How do you change the voice and the speed?

The voice and the speed live in different places, which is why people think they can't be changed.

Voice. In System Settings, Accessibility, Spoken Content (or Read & Speak), use the System voice menu to pick a voice, and System speech language to change the language. Apple's page for this ("Change the voice your Mac uses to speak text") says to click the Info button next to System voice to see more voices, click a voice's name, and press Download if a download button is shown. Apple says the voice is downloaded from Apple, you must be connected to the internet, and "voices aren't available until the download is complete." Not every language has voices.

Apple's pages don't rank the voices by quality, so the only reliable method is to press the Play button beside a name and listen. Give each a full sentence, not a word. A voice that sounds fine on "Hello" can wear thin by the fourth paragraph.

Speed. Two places. While the Mac is speaking, the controller changes the speaking rate (one button on Tahoe 26, separate decrease and increase buttons on Sequoia 15), which is the easier route because you hear the result immediately. In settings, the Info button next to System voice lets you adjust "how fast or loud it speaks," in Apple's words, and Play gives you a sample.

One thing to try that Apple's pages don't cover: if a voice gets tiring after ten minutes, slow it down a notch before swapping voices. That's a hunch from listening, not a documented fix.

What are the other built-in options, and which one is for what?

There are five more tools in the same neighbourhood. They solve different problems, and mixing them up is a common way to decide the Mac "can't do it."

A few notes on that table, because each row has a catch.

Speak items under the pointer. Apple says you choose whether it speaks all the time or only when you use zoom, and you can set verbosity (high, medium or low) and a delay before it speaks. It's about interface items, not article text.

Live Speech speaks what you type, such as in-person conversations or FaceTime. Apple's steps: Accessibility, Speech, Live Speech, turn it on, type in the Type to Speak field and press Return. Apple notes it isn't available in all languages. If you've landed on it while hunting for "Mac text to speech," it's the wrong tool.

Accessibility Reader is the newest of these. Apple's page describes it as showing text from any app in a full-screen view where you can change fonts, spacing and colours, and listen with play, pause, skip, speed and scrubbing controls. Default shortcut is Command-Escape, switched on under Read & Speak. Apple's page I read is the one for the newest macOS, and it doesn't say which version introduced the feature, so check that it's in your Read & Speak pane before counting on it.

VoiceOver is a screen reader built for people who can't see the screen. Apple's page puts it plainly: press Command-F5 to turn it on, and its Read All command (VO-A, where VO is Control and Option together, or Caps Lock) reads from the top of the page. It will read your article. It will also narrate every menu and button on the way, because that's its job. For casual article listening it's too much machine.

Safari Reader has no audio feature at all. Apple's page for it, "Hide distractions when reading articles in Safari on Mac," describes a view of the article on one page without navigation or distractions, and says the option only appears when Safari finds an article it can show. It doesn't mention listening. Apple doesn't document using it together with Speak selection. But the two descriptions fit together: Reader removes the navigation and cookie banner, and Speak selection reads the available text in the window. If the Mac keeps reading menus before it gets to the article, that pairing is the first thing I'd try.

Why do the built-in voices stop working on long articles?

The built-in voices suit a paragraph and fit a sitting badly.

Here's what the Apple pages describe, and what they don't. They describe highlighting for words and sentences, an onscreen controller, and a window-level read when nothing is selected. They don't describe a saved position, a library of things you've been listening to, or a way to pick up tomorrow where you stopped today. None of the controller's buttons is "resume where I left off." That absence is the session gap in concrete form.

Four things follow from it.

You lose your place. Stop to answer a message and, as far as Apple's pages describe, nothing takes you back to the sentence you'd reached. On a 400-word email that's nothing. On a 5,000-word article it's a reason to abandon it.

You hear the page, not the article. With nothing selected, the Mac reads available text in the window. That includes whatever else is in the window. Selecting the article body by hand fixes it, and Safari Reader may help.

One voice setting covers the lot. The voice comes from the System voice menu, and Apple's pages describe no per-article choice. A voice that's fine for a sentence isn't automatically one you'd choose for an hour of long-form reading.

Nothing connects what you heard to what you keep. This is the one that matters most, and it isn't really about Apple.

Take the research. Rogowsky, Calhoun and Tallal published a study in SAGE Open in 2016 in which 91 participants got the same non-fiction material (a preface and a chapter of a book) as a digital audiobook, as an e-text, or both together. There were no statistically significant differences in comprehension, either right after or two weeks later. It's one study, one text, and I wouldn't build a theory on it, but it points somewhere interesting. The choice between listening and reading is not what decides what you remember.

So if the Mac's voice does read the article to you, you may well have done much the same as reading it. And you've also done nothing that makes it stay. Listening, like reading, is a way of getting words in. Whether anything is still there in two weeks is a separate question, and no setting in System Settings answers it.

That's the relief in all this. You haven't been failing at text-to-speech. You've been asking a one-shot tool to hold a sitting, and it was never going to remember the sitting for you.

When does a dedicated reader help?

I didn't start building Alexandria because Mac text-to-speech was bad. I started because of a Sunday in May 2025. I'd come home from the gym and sat down with client documents I'd been avoiding, and the words weren't going in. I wondered what would happen if the computer read with me instead of to me.

That phrase did a lot of work. "To you" is Speak selection. "With you" means the voice and the words moving together, a place that's held when you step away, and a speed you can set without leaving the page.

Here's how that looks in Alexandria, using only what it does today. It's a Chrome extension (desktop) plus a web app, so it reads in Chrome on your Mac, and the web app is where PDFs go. It highlights the word being spoken and the sentence it belongs to at the same time, so your eye always knows where the voice is. Neural voices are on the free tier, and free speed runs from 0.5x to 5x. Your library, highlights and reading position sync across devices. It never stores the text it reads aloud: the audio cache is keyed on a one-way hash of the text, with no column anywhere that holds your words. It adds a play button directly into places like Gmail and Reddit. PDFs are read in the web app's reader.

What it won't do: read a Pages document, a native Mail message or anything outside Chrome and the web app. The built-in tools can. If most of your reading is in native apps, stay with Speak selection.

If you want to compare Alexandria against other extensions, the Chrome extension comparison lays out voices, speed and privacy side by side. For PDFs specifically, how to listen to a PDF walks through the Mac route and the alternatives, and listening to articles shows the web version.

I've put that last row in on purpose. A dedicated reader that only plays audio has the same ceiling as the Mac. What I wanted was for the time I spent listening to stop evaporating. If you'd like a way to do that without changing tools, how to remember what you read is the method, and it works with Speak selection too.

What should you try first?

Start with the two-minute version.

Turn on Speak selection, pick a voice you can stand by pressing Play on a few, switch word and sentence highlighting on, and open the article in Safari Reader. Press Option-Esc. If it holds up for ten minutes, you may never need anything else, and that's a fine outcome.

If you catch yourself scrolling to find where the voice got to, or you're dreading the 5,000-word article before you've started it, that's the session gap, and now it has a name.

Frequently Asked Questions

What is the shortcut to make a Mac read text aloud?

The default shortcut is Option-Esc, once Speak selection is switched on in System Settings under Accessibility. Select some text and press it to start, then press it again to stop. You can change the shortcut from the Info button next to Speak selection.

Is it Spoken Content or Read & Speak?

It's the same pane under two names. Apple's pages for macOS Sequoia 15 call it Spoken Content, and its pages for macOS Tahoe 26 and later call it Read & Speak. Both live under Accessibility in System Settings, and Speak selection sits inside either one.

How do I make my Mac read faster or slower?

Use the onscreen controller while the Mac is speaking, which lets you change the speaking rate. You can also click the Info button next to System voice in System Settings and adjust how fast the voice speaks. The controller is easier mid-article, because you hear the change straight away.

How do I download a better voice on a Mac?

Open System Settings, Accessibility, then Spoken Content or Read & Speak. Click the Info button next to System voice, click a voice's name, and press Download if it's shown. Apple says the voice is downloaded from Apple, you must be online, and voices aren't available until the download is complete.

Can a Mac read a whole web page aloud?

Yes. With nothing selected, Speak selection reads the available text items in the current window, according to Apple. That covers a web page open in your browser. Apple's pages describe no way to save your place, so stopping partway means finding your spot again.

Does Live Speech read articles aloud?

No. Live Speech speaks text you type, for example in conversation or on a FaceTime call. It lives under Accessibility, then Speech, in System Settings. To hear an article you already have on screen, use Speak selection, or paste the article into the Live Speech field if you're stuck.

When is a dedicated text-to-speech app worth it on a Mac?

It's worth it when you listen in long sessions, return to the same material, or want voices and speed beyond the system defaults. For a paragraph or an email, the built-in tools are enough. The gap opens with 20-minute articles, where saving your place and retention start to matter.

---

Related reading: Best Text-to-Speech Chrome Extensions | How to Listen to a PDF | Listen to Articles | How to Remember What You Read