Google has begun rolling out advanced voice control capabilities for Gemini on macOS, first previewed at the I/O 2026 conference in May. The update promises to transform how users interact with their desktop, offering intelligent dictation and context-aware commands based on the screen content. With version 1.88 of the app, users can now speak naturally into any window, with accuracy comparable to Gboard Rambler on upcoming Gemini Intelligence smartphones.
Intelligent dictation polishes speech in real time
The first major feature is advanced dictation, which transcribes spoken words into clean, polished text. Gemini automatically removes fillers like “umm” and “ah” and handles mid-sentence corrections, capturing the gist of what you say. The formatted text is entered at the current cursor position, offering a speech-to-text experience similar to Gboard Rambler. This is especially useful for composing emails, documents, or messages without using the keyboard, boosting productivity for Mac users. Unlike other dictation systems, Google’s AI understands sentence context and adapts punctuation naturally, without requiring specific voice commands.
Sponsored Protocol
Beyond transcription: understanding the screen to execute tasks
The second capability goes beyond transcription: Gemini can interpret the content of your screen to perform complex tasks. For example, if you are viewing a map, you can ask to find a nearby restaurant and add an appointment to your calendar, all with a single voice command. This feature is opt-in and requires enabling “Gemini reasoning” in the app settings. To activate, simply long-press the Fn key anywhere in macOS, or tap the new screen sharing button at the end of the “Ask Gemini” prompt bar. A floating pill with a waveform appears at the bottom of the screen, indicating Gemini is listening. Google notes that this capability is still rolling out globally to English-speaking users, with more languages coming soon.
Sponsored Protocol
Availability and getting started
To take advantage of the new features, users must update Gemini for macOS to version 1.88, available on the Mac App Store or the official website. The rollout is gradual, so it may take a few days for everyone to receive the update. This move marks a significant step for Google in competing with built-in voice assistants like Siri and Cortana, bringing Gemini’s power directly to the desktop. AI security remains a hot topic: as highlighted by a recent FAR.AI test, some large language models show vulnerabilities to jailbreaks, while Gemini itself proved immune to automated attacks. Meanwhile, the Mac lineup is bracing for new models: the Mac Mini M5 reports wait times of up to three months and a 25% price increase. For more background on the underlying technology, check the Wikipedia page on Google Gemini.
Sponsored Protocol
Source: https://9to5google.com/2026/07/29/gemini-mac-voice-control
