Gemini for Mac adds voice input across any app
By Steven Van ·
Long-pressing Fn transcribes and cleans up speech anywhere on macOS, with context-aware commands to summarize, rewrite, or edit at the cursor.
Gemini's macOS app now supports voice, letting you create, edit, and summarize by speaking into whatever window you're already working in rather than switching to a chat interface first. The feature works anywhere on the desktop, which is the detail that separates it from a typical voice-to-text add-on scoped to a single app.
Intelligent dictation, not just transcription
Long-pressing the Fn key activates voice input that transcribes speech into polished text, automatically stripping filler words and correcting formatting along the way. That's a meaningfully different result than raw dictation, which usually needs a manual cleanup pass, umms and false starts included, before it's usable.
Context-aware commands, not just text entry
With context-aware reasoning enabled in settings, Gemini can understand what's on screen and act on it by voice: extract and summarize information from open files and documents, rewrite or adjust the tone of highlighted text, or generate and edit images from a spoken description. Results insert directly at the cursor position, so the workflow stays inside whatever app you're already in rather than bouncing to a separate Gemini window and copying the result back.
Global rollout, English first
The feature is rolling out globally to all Gemini app users on macOS starting today, initially in English with more languages coming soon. It's available now at gemini.google/mac.
Part of Gemini's push into ambient assistance
Voice-anywhere input fits the same direction as Gemini Spark's new Chrome integration, both aim to make Gemini something that works alongside whatever you're doing rather than a destination you have to open and switch into.
Sources: Google, Gemini on X.