Gemini launches Mac global voice input: Edit directly by speaking
According to on-chain analytics firm Beating, Google has rolled out global voice typing for Gemini on macOS. Users can simply long-press the Fn key in any app to start speaking, and the transcription will be inserted directly at the current cursor position. Gemini automatically removes fillers such as "um" and "uh," recognizes mid-sentence corrections, and organizes the content into paragraphs or lists.
With optional on-device inference enabled, Gemini can also read selected text, files, images, and documents. Users can verbally instruct it to paraphrase content, adjust tone, summarize, or generate and modify images, with the results being written back to the current app.
This new feature is gradually rolling out globally to all Gemini Mac users, with initial support available only in English.