Gemini Speak to Window Explained: Talk to Your Mac in Any App
Google switched on Speak to Window in the Gemini app for macOS on July 29, 2026. Long-press the Fn key, talk, and clean text lands at your cursor in whatever app you are in. It is free and needs no subscription, but the app runs only on Apple Silicon MacBooks with macOS Sequoia 15 or later.
Google just made your Mac's keyboard optional. On July 29, 2026, Google switched on a new voice feature in the Gemini app for macOS that works in any window: hold down the Fn key, say what you want, and polished text appears at your cursor in whatever app you happen to be in. Google calls it Speak to Window. The important consumer detail is not the price, because it is free -- it is the hardware. The Gemini Mac app only runs on Apple Silicon MacBooks with macOS Sequoia or later, so whether you get this comes down to which Mac is on your desk.
Key Takeaways
- It is free and needs no subscription. Google's release note says the feature is "rolling out today in English to all users on the Gemini macOS app," and the app itself is "available at no cost" everywhere Gemini is supported, per Google's Mac page.
- The real gate is your Mac, not your plan. Google says the app "runs exclusively on Apple Silicon MacBooks and requires macOS Sequoia (15.0) or later."
- Hold Fn and talk. Long-press the Fn key (the one marked with a globe), speak, then release to submit. Double-tap Fn instead for longer hands-free dictation.
- The typing half works out of the box. Intelligent dictation is on by default: it strips out "ums" and "ahs", catches your mid-sentence corrections, and formats what you said into tidy paragraphs or bullets.
- The clever half is opt-in. Turn on reasoning under Settings, then Speak to Window, and Gemini can read your screen to rewrite highlighted text, summarise files you have selected, or generate an image where you are working.
- English only for now, with Google saying more languages are "coming soon" and giving no date.
- It lands in your app, not in a chat window. That is the actual difference from opening Gemini and pasting the result back.
What Is Speak to Window?
Speak to Window is a system-wide voice layer that types for you in every Mac app, not a chatbot you visit. You hold a key, speak, and the finished text is dropped straight into the document, email, message box, or form you already had open. Nothing switches to a Gemini window and nothing needs pasting back. Google's framing is that the Mac app "was built to keep you in your flow," and this is the feature that most obviously delivers that.
In everyday terms: it is voice typing that thinks a little. Ordinary dictation writes down exactly what leaves your mouth, filler and false starts included. Google says this one automatically removes the "ums" and "ahs," catches mid-sentence corrections, and formats the result, so what appears is closer to what you meant than to what you said.
Do You Get It? Check Your Mac First
This is the unusual part of this release: the answer depends on your hardware, not your wallet. There is no paid tier for Speak to Window and no upgrade prompt. But the Gemini Mac app has hard requirements, and if your machine misses them there is no version of this you can install.
| Your setup | Do you get it? | Why |
|---|---|---|
| Apple Silicon MacBook, macOS Sequoia 15.0 or later | Yes, free | Meets Google's stated requirements for the app; the feature is going to all users of it |
| Apple Silicon MacBook, older macOS | No, until you update | Google requires Sequoia (15.0) or later |
| Intel Mac | No | Google says the app "runs exclusively on Apple Silicon" |
| iMac, Mac mini, Mac Studio, Mac Pro | Check before you plan around it | Google's own wording says "Apple Silicon MacBooks," and it publishes no separate desktop line |
| Windows PC or Chromebook | No | This is a macOS app feature |
| iPhone, iPad, Android | No | The Gemini mobile apps are not part of this rollout |
| Any of the above, but not in English | Not yet | Rolling out first in English; more languages "coming soon," no date |
The plan question has a short answer: it does not matter. Free Gemini accounts get this exactly as Google AI Pro and Ultra subscribers do. That is worth stressing because Google's other headline Mac feature, Gemini Spark, is the opposite -- a paid, region-restricted agent. Speak to Window has none of those strings.
What Works by Default, and What You Have to Switch On
Two different things live behind the same key, and only one is active when you install the app. Google separates them clearly, and it is worth knowing which you are using before you wonder why Gemini is not doing more.
| Capability | On by default? | What it does |
|---|---|---|
| Intelligent dictation | Yes | Transcribes you into clean text at your cursor, removing filler words, applying your mid-sentence corrections, and formatting into paragraphs or bullet points |
| Reasoning with screen context | No, opt-in | Lets Gemini act on what is on screen: rewrite or reformat highlighted text, pull details out of selected files, images, or PDFs, and generate images in place |
To turn the second one on, Google says to go to Settings, then the Speak to Window tab, and enable reasoning. After that, Gemini "intelligently distinguishes between dictation and reasoning tasks," so you do not have to tell it which mode you want -- you just talk.
That opt-in is a sensible default rather than a hoop. Reasoning means Gemini reads the content of your active window, and Google has left that switched off until you decide you want it. If you only ever want voice typing, you never have to turn it on.
How to Actually Use It
Google's instructions are short, so here they are in order:
- Download the app from gemini.google/mac and sign in.
- Put your cursor where you want the text, in any app.
- Press and hold the Fn key (marked with a globe on most Mac keyboards) and start speaking.
- Release the key to submit. Gemini drops its response into your active window.
- For longer dictation, double-tap Fn instead to start hands-free recording, then tap again to finish. You can also click the Speak to Window icon in the prompt bar.
There is one genuinely useful trick in Google's own tips. To point Gemini at something specific, highlight the text or image first, then long-press Fn and speak. Google adds a warning worth heeding: deselect the highlighted content after you start speaking, or Gemini will replace it.
If the Fn key is already spoken for on your Mac, you are not stuck. Apple's Mac User Guide lists "Press Fn (Function) Key Twice" as one of the shortcut options for macOS's own Dictation, so a collision is plausible if you use both. Google says you can customise the shortcut keys under Settings, then Speak to Window.
How This Compares to Dictation You Already Have
Most people reading this already have at least two ways to talk instead of type. The differences are real but narrow, and they come down to where the text ends up and how much cleaning happens on the way.
| Option | Where the text lands | What is different |
|---|---|---|
| Gemini Speak to Window | Directly at your cursor, in any Mac app | Cleans filler and formats automatically; can act on screen content once you opt in |
| macOS Dictation | At your insertion point, in any Mac app | Built into the operating system, no download; started with the Microphone key, a keyboard shortcut, or Edit, then Start Dictation |
| ChatGPT dictation | In ChatGPT's own message box | Got a more accurate speech-to-text model in June 2026, on every plan including free -- but you are typing into ChatGPT, not into your document |
| GPT-Live and voice modes | Nowhere; it is a spoken conversation | A back-and-forth chat where the assistant talks to you, not a way to produce text in your own files |
Put plainly: ChatGPT's dictation makes talking to ChatGPT better. Gemini's Speak to Window makes talking to your Mac better. They solve different problems, and if you write in Pages, Mail, Slack, or a browser text box all day, the second one saves more steps.
Should You Care?
Get it if you have an Apple Silicon MacBook, write a lot, and have ever given up on voice typing because cleaning up the transcript took longer than typing would have. That is the specific complaint this is aimed at, and it costs nothing to test.
Skip it if you are on an Intel Mac or an older macOS, in which case there is no decision to make, or if you do not write in English yet. Do not upgrade any plan for this -- paying Google more does not get you a better version of it.
Leave reasoning off if the idea of an assistant reading your active window bothers you. Dictation alone still works, and Google keeps the screen-reading half behind a switch you control.
The honest verdict: this is a small feature with an unusually clean answer for once. It is free, it is global in English, it needs no subscription, and the only thing standing between most people and it is whether their Mac is new enough. That is rarer than it sounds -- if you are trying to work out which assistant is worth paying for at all, our guides to which AI chatbot you should pay for and ChatGPT vs Gemini for everyday use cover the parts that do cost money.
The bottom line: Gemini's Speak to Window turns your Mac's Fn key into a dictation button that works everywhere and tidies up after you, for free. Check that you have an Apple Silicon MacBook on macOS Sequoia or later, download the app, and hold down Fn. If you want it to do more than type, switch reasoning on yourself.
Frequently Asked Questions
Is Gemini's voice feature on Mac free?
Yes. Google's release note says it is rolling out to all users of the Gemini macOS app, and Google describes that app as available at no cost everywhere the Gemini app is supported. There is no plan to buy and no upgrade prompt for this feature.
Which Macs can run the Gemini app?
Apple Silicon MacBooks running macOS Sequoia (15.0) or later, in Google's own words on its Mac page. If your Mac has an Intel processor or an older version of macOS, you cannot install the app, so this feature does not reach you at all.
How do I start talking to Gemini on my Mac?
Press and hold the Fn key, also marked with a globe, and start speaking. Release it to submit. For longer stretches, double-tap Fn to record hands-free and tap again to finish. You can also click the Speak to Window icon in the prompt bar.
Does it just type, or can it do more?
Both, but only one is on by default. Intelligent dictation, which cleans up and formats your speech, works out of the box. Reasoning tasks that use your screen, like rewriting highlighted text or making images, are opt-in under Settings, then Speak to Window.
Is this available in my language?
Only English at launch. Google says the feature is rolling out first in English to all users in the Gemini macOS app, with more languages coming soon. It has not published a list of which languages are next or when they arrive.
Will it clash with my Mac's built-in dictation?
It might, since one of Apple's own Dictation Shortcut options is pressing the Fn key twice. Google says you can customise Speak to Window's shortcut keys in the app's settings, so if the two collide you can move Gemini onto a different key.