Google has released version 1.88 of its Gemini desktop application for macOS. The update, which began rolling out on July 29, introduces voice input capabilities that function across active windows and a new reasoning mode for analyzing on-screen content.
Users can activate voice input by long-pressing the Fn key or using custom shortcuts. The system transcribes speech directly into applications, automatically removing filler words and performing corrections. A floating waveform indicator shows when the tool is active.
The reasoning mode allows Gemini to process documents, images, and spreadsheets directly on the screen to perform tasks like summarization and data extraction. This feature is disabled by default and requires manual activation in the app settings.
The release follows a preview of these features at Google I/O in May 2026. Currently, the update is available globally but supports only the English language. Google has not yet announced a timeline for adding support for other languages.
