Overview
- Google rolled out the new voice features for the Gemini macOS app on July 29, making intelligent dictation and optional screen awareness available globally in English.
- Intelligent dictation is enabled by default and is triggered by a long press of the Fn key, transcribing spoken words into cleaned, formatted text that drops directly at the cursor.
- An opt‑in screen‑aware reasoning mode lets Gemini read highlighted on‑screen content and local files to extract, summarize, rewrite, generate images, or synthesize across multiple files without manual copy‑paste.
- The app runs via users’ Google accounts and processes requests in the cloud under Google’s Gemini Apps Privacy Policy, which allows data retention and potential human review and raises concerns about prompt‑injection and other security risks.
- The update extends capabilities first shown in April and in the June Gemini Spark beta, positioning Google to compete with Apple Intelligence and third‑party desktop assistants and signaling faster voice‑driven workflows on the Mac.