Google adds voice AI to 'Gemini' for macOS, letting users control AI directly by speaking during tasks

Google has rolled out the voice AI feature “Speak to Window” for the Gemini desktop app on macOS.

Using the Fn key and voice alone, users can type, edit documents, summarize, rewrite, and generate or edit images inside the app they are currently working in.

The feature is available in English first and requires Apple Silicon Macs running macOS Sequoia 15 or later.

It expands Gemini beyond a standalone chatbot into an OS-wide work tool, sharpening Google’s competition with OpenAI and Anthropic.