Gemini for Mac gets smarter voice controls for writing, editing, and summarizing content

Google is rolling out new voice capabilities for the Gemini app on macOS, allowing users to dictate text and control AI directly inside the active application without switching windows.

The feature is activated by holding the Fn key and speaking. By default, Gemini works as an AI-powered dictation tool. It removes filler words and pauses, adapts to corrections made while speaking, formats the text, and inserts the final result wherever the cursor is placed.

What can Gemini do?

When Reasoning Mode is enabled, Gemini can also understand the contents of the active window, making it possible to perform more advanced tasks with voice commands.

For example, users can ask Gemini to:

• summarize selected documents, images, or local files;
• rewrite highlighted text, adjust its tone, or turn notes into a structured document;
• generate new images or edit an existing one.

Google says users can select several documents and ask Gemini to draft an email based on their contents. Another example is converting raw notes into a structured summary with key takeaways, with the result inserted directly into the active application.

For longer dictation sessions, users can double-press the Fn key instead of holding it down continuously. The feature can also be launched from the Speak to Window button in Gemini's menu bar app, while keyboard shortcuts can be customized in the app's settings.

Availability

The new features are rolling out globally to Gemini users on macOS, but are currently available only in English. Google says support for additional languages will arrive later.

The Gemini app requires a Mac with Apple Silicon running macOS 15 Sequoia or later.