Google Gemini Adds Global Voice Input for macOS with On-Device Inference

Gemini introduces voice typing on Mac, allowing users to edit text via speech with automatic formatting and on-device processing for privacy.

Woofun AI reports that Google has deployed global voice typing for Gemini on macOS, enabling users to initiate transcription by long-pressing the Fn key in any application. The system inserts text directly at the cursor, automatically filtering filler words and structuring content into paragraphs or lists based on verbal cues.

With optional on-device inference, Gemini can process selected text, files, images, and documents locally. Users may issue voice commands to paraphrase, adjust tone, summarize, or modify images, with results integrated back into the active application. The feature is rolling out globally to all Gemini Mac users, currently supporting English only.