Google has added voice-based capabilities to its Gemini app for macOS, allowing users to dictate, edit, and summarize content across desktop applications using their voice.

The feature is activated by long-pressing the Fn key, which opens a voice input mode within any open window. By default, the tool operates as an intelligent dictation system, transcribing spoken words into formatted text while removing filler words and mid-sentence corrections.

Users who enable an optional "Gemini reasoning" setting gain access to additional functions. These include the ability to extract and summarize content from local files, images, or documents; rewrite or adjust the tone of highlighted text; and generate or edit images based on voice instructions referencing on-screen content.

The capabilities are rolling out globally to all Gemini for macOS users in English, with support for additional languages described as coming at a later date.