Google’s Gemini Mac App Just Learned How to Listen
Toggle Dark Mode
As Apple prepares to launch Siri AI as part of macOS 27 Golden Gate this year, Google is continuing to up the stakes with its own Gemini app for Mac. Earlier this month, the company rolled out an agentic AI tool for automating desktop workflows, and now it’s pivoting away from the text-based chatbot to offer native voice assistance and transcription.
Google announced the new capabilities today, although they’re not entirely encapsulated in an app update, but rather appear to be slowly rolling out from the back end. So, even if you have the latest version of Gemini for Mac installed (1.88.5.636 as of this writing), you won’t necessarily see the new features right away.
Once they are up and running, however, Google says that you’ll be able to “create, edit, and summarize with Gemini using just your voice” — including dictating directly into virtually any document or text field.
By long-pressing the Fn key, you can speak naturally into any window on your desktop. By default, this enables intelligent dictation, letting you easily transcribe your spoken words into clean, polished text — automatically removing the “ums” and “ahs,” catching mid-sentence corrections, and dropping the formatted text instantly at your cursor.
This mode will bypass the standard Gemini window, inserting text directly into whatever app you’re using.
That’s not a new feature for Mac users — Apple has provided a built-in dictation feature for years, with recent improvements in macOS Sonoma allowing for simultaneous talking and typing without having to switch between the two. We’ll have to wait and see how Gemini’s version compares to what’s already here, but even if it beats out macOS Tahoe, Apple already has a big upgrade in the pipeline for macOS 27 Golden Gate, thanks to Siri AI and more advanced on-device models — provided your Mac has the chops to support them.
During its 2026 Worldwide Developers Conference (WWDC), Apple noted that the advanced models will require a device with 12 GB of RAM. That encompasses nearly all of the Macs Apple currently sells, but there may still be more than a few early Apple silicon Macs with only 8 GB.
In a similar vein, the new update will allow users to enable Gemini reasoning through voice prompts, adding support for screen context. For example, you could highlight a set of files and ask Gemini to look up specific information in them (e.g. “Read these vet files and summarize my dog’s medical history in an email to the kennel.”), highlight text in a document or email and ask Gemini to rewrite it, or generate or edit images.
Google has been rapidly improving Gemini for Mac since its launch in April, bringing the promised “Desktop Intelligence” earlier this month, and now rolling in voice control. Many of these are things that Apple is promising for Siri AI in macOS Golden Gate, but this Gemini could still be a bonus for folks who want to stay on an older macOS version or simply prefer to work in the Google AI ecosystem.
Gemini for Mac is a free download from Google’s website, and is available in all languages and countries where the Gemini app is supported. It’s compatible with macOS Sequoia 15.0 or later, but still requires Apple silicon, which means every Mac capable of running Gemini natively will also be eligible to run macOS Golden Gate later this year with most, if not all, of the Siri AI capabilities that Apple is promising.
