Gemini macOS App Launches Global Voice Feature 'Speak to Window'

Google officially announced the global rollout of 'Speak to Window' for the macOS Gemini desktop app. Users can hold the Fn key to invoke voice input in any active window, enabling dictation or screen-context reasoning. This feature breaks the limitation of a single app window, seamlessly integrating the AI assistant into the desktop workflow, and is now available to English users worldwide.

Confirmed

How it works: Users can hold the Fn key to activate voice input. Additionally, the Gemini app can be summoned anytime via the Option + Space shortcut.
Context reasoning: The app intelligently distinguishes between standard dictation and screen-context reasoning tasks. To enable screen-context reasoning, users must manually turn it on in the app's 'Speak to Window' settings.
Supported tasks: The feature supports extracting and summarizing information, writing and rewriting text, and generating and editing images. Gemini can also convert highlighted text, analyze local files, and generate visual charts directly at the cursor.
Availability: The feature is now available globally, primarily supporting English interaction.

Why it matters

This feature enables cross-app global voice operation without switching to a separate conversation window, deeply integrating Gemini's multimodal capabilities into users' daily desktop workflows, significantly enhancing local operation experience and AI productivity.

2026-07-30 ~ 2026-08-01 · 7 related posts

Primary sources

Gemini macOS App Gets Voice Commands for Global Dictation and Editing — GeminiApp · 2026-07-31 00:00 (UTC+08:00)
Gemini macOS App Launches Voice Interaction: Press FN to Dictate and Reason — GeminiApp · 2026-07-31 00:00 (UTC+08:00)
Gemini macOS App Update: Long-Press FN to Activate Screen Context Reasoning — GeminiApp · 2026-07-31 00:00 (UTC+08:00)

Gemini macOS App Gets "Speak to Window": Long-Press Fn for Voice UI — testingcatalog · 2026-07-30
[source] Gemini macOS App Gets Voice Commands for Global Dictation and Editing — GeminiApp · 2026-07-31
[source] Gemini macOS App Update: Long-Press FN to Activate Screen Context Reasoning — GeminiApp · 2026-07-31
Gemini App Launches on macOS with Natural Voice Interaction — emmanuelvivier · 2026-07-31
Google Launches Native Gemini Mac App with Voice and Screen Automation — GeminiApp · 2026-07-31

2 near-duplicate retellings: GeminiApp · testingcatalog