Gemini's July update adds voice control on Mac, and leaves Europe out of the headline feature
The July Gemini Drop brings dictation into any window on macOS, new Flash models, app connections and personalised images. Gemini Spark goes worldwide, except in the EEA, the UK and Switzerland.
Google has published the July edition of Gemini Drops, its monthly summary of what changed in the Gemini app. Six items this time, and a footnote that will matter more to European readers than anything above it.
The most immediately useful addition is voice control on macOS. You can speak into any active window to dictate text, rewrite whatever you have highlighted, or generate an image where your cursor sits. That is a meaningful shift in where the AI lives: instead of switching to a chat tab, copying text over, and copying the result back, the model comes to the document you are already in. Two new models are also live, Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, which Google describes as faster with better reasoning. Flash models are the cheap, quick tier meant for high-volume everyday work rather than hard problems.
The rest is smaller. Gemini can now connect to Dropbox, Zillow Rentals and Viator, so it can act across those services from a single prompt. You can save an avatar and have Gemini put you into generated images without re-uploading a photo each time. Deeply personalised image generation, tuned to your stated interests, is rolling out to all users in the US. And Gemini Spark, Google’s agent that keeps working after you close the laptop, is going worldwide. Except, per Google’s own footnote, in the European Economic Area, the United Kingdom, Switzerland and Nigeria.
What is behind this. That footnote is the story for anyone reading from Germany, Austria or Switzerland, and it is becoming routine. Agentic features that act on your behalf across services touch data protection and the AI Act in ways a chat window does not, so they now ship to the rest of the world first and arrive in Europe months later, if at all. This is not Google singling Europe out; OpenAI and Meta have made the same call repeatedly. The trade is real and worth naming honestly rather than complaining about: stronger rules, slower access. Reasonable people weigh that differently.
What this means for you: if you use a Mac and Gemini, turn on the voice feature and give it a week, because dictating edits directly into the window you are working in genuinely removes friction. Everyone can switch to the new Flash models for routine tasks. If you are in the EU, the UK or Switzerland, do not go hunting for Gemini Spark, it is not there, and the personalised image feature is US-only for now. For anyone comparing assistants, the pattern to watch is that the interesting releases increasingly happen at the operating system level rather than inside a chat app, and that is where the practical differences between these tools will be decided.
Sources
Source: https://blog.google/products-and-platforms/products/gemini/gemini-drop-july-2026/
Google DeepMind's new robot brain is meant to run everything from a desk arm to a humanoid
Gemini Robotics 2 and Gemini Robotics ER 2 split the job of controlling a robot into thinking and moving. Here is what a vision-language-action model actually is, and what is still on a waitlist.
