Gemini for macOS adds new natural language capabilities

Google is rolling out new natural language capabilities for the Gemini app on macOS, designed to let users create, edit, and summarize content using only their voice while working in any desktop window.

By long-pressing the function key, users can activate intelligent dictation that automatically transcribes spoken words into polished text by removing filler words and capturing mid-sentence corrections, placing the final output directly at the cursor.

Users can also enable Gemini reasoning in the settings to execute more complex tasks based on the context of their screen. This allows people to extract and summarize information from highlighted local files, images, and documents, compose and rewrite highlighted text by adjusting the tone, and generate or edit images by referencing desktop designs with spoken commands.

The updated voice functionality is currently rolling out globally in English for all macOS Gemini app users, with plans to add more languages in the future.