Gemini adds voice on macOS and new AI features
Google introduces its July 2026 Gemini updates: voice integration on macOS, new Flash models, connections to Dropbox, Zillow Rentals, and Viator, and tools for creating personalized images. Gemini Spark is also going worldwide, while some features remain limited to the United States.
Gemini can now work with your voice on macOS, and Google is expanding its assistant with new models, connections to other apps, and tools for creating more personal images. The updates are part of the July 2026 Gemini Drop.
Gemini works inside your apps
The Gemini app for macOS lets you talk to it from any active window. You can dictate polished text, transform a passage you have selected, or generate an image without switching apps.
In practice, you could select a paragraph and ask Gemini to make it clearer, dictate an email reply, or create a visual asset while working in another program. Everything happens from the window you are already using.
Google is also expanding actions connected to other apps. Gemini can work with Dropbox, Zillow Rentals, and Viator when you link those accounts and give it an instruction from one place.
That lets you, for example, search information stored in Dropbox, browse rentals, or plan travel activities without opening each service separately. The usefulness will depend on the permissions and capabilities of each connected app.
New Flash models
Google is adding Gemini 3.6 Flash and Gemini 3.5 Flash-Lite. The company describes them as its latest Flash models, with improvements in reasoning and speed.
The difference matters most when you want fast answers or need Gemini to follow several steps to complete a task. The announcement does not provide specific performance figures in this update or compare the models with earlier versions.
An avatar for creating images
Gemini also lets you add yourself to images through an avatar. Once you create one, you can generate content that looks like you without uploading a new photo every time.
In addition, all users in the United States can now create images tailored to their interests and preferences. Google describes this as a way to produce more personalized images, although the feature has not yet been announced as available in every country.
Gemini Spark goes worldwide
Gemini Spark, a feature designed to run tasks even after you close your laptop, is now available worldwide. Its goal is to let the assistant keep helping you throughout the day instead of limiting it to the session you have open.
These updates point to a less isolated Gemini: it can understand what is in front of you on macOS, connect to services you already use, generate content with a more personal visual identity, and rely on new models. What matters now is seeing which features reach each country and how much control you retain over your data when connecting apps and avatars.