Gemini for macOS adds natural voice features

Gemini for macOS now lets you dictate, summarize, rewrite text, and edit images using only your voice. The feature activates when you hold down the Fn key while working in any desktop window.

By default, Gemini works as an intelligent dictation system. It transcribes what you say, removes pauses and filler words such as “uh” or “um,” understands corrections you make halfway through a sentence, and places the polished text directly where your cursor is.

You do not need to open a separate app or copy and paste the result. You speak into the document, email, or text field you are already working in.

It can also act on what is on your screen

You can enable Gemini’s reasoning in the settings. With this option, the assistant can interpret the context on your screen and help with more involved tasks, using only the files, images, or text you select.

For example, you can highlight several local documents and ask it to:

• Extract the important information and summarize it.
• Draft an email based on that information.
• Rewrite text in a different tone.
• Turn notes into an executive summary with a synthesis at the beginning.
• Generate or modify an image using an illustration you have open as a reference.

A prompt such as “read these files from the vet and summarize my dog’s medical history in an email for the boarding facility” combines several tasks that would normally require reviewing documents, drafting, and editing separately.

You can also request quick changes to an image. For example, you can select an illustration and ask for a dark mode version without having to describe it from scratch.

What changes for you

The update reduces the number of steps between an idea and the result. You can dictate a message in a specific tone, summarize several files without switching windows, or modify an image while reviewing it.

The important difference is that Gemini is not limited to converting speech into text. When you enable reasoning, it uses the content you have selected and the visible context to carry out a more complete task.

The feature is rolling out globally to all Gemini users on macOS, for now only in English. Google says it will add more languages soon. To try it, you need to download the app from gemini.google/mac.

The next step will depend on the available languages and how accurately Gemini interprets spoken instructions and local content. For now, the key point is simple: on macOS, holding down Fn turns your voice into a direct way to write and work with what is in front of you.