Google is introducing Gemini in all cars that support Android Auto in the coming months. The company announced it at the Android show before its 2025 I/O Developer Conference.
The company has added Gemini functionality to Android Auto, and in a blog post it will drive “more productive and fun” to cars running Google’s built-in operating system later this year.
“I think this will be one of the biggest changes in the vehicle experience we’ve seen for a very long time,” said Patrick Brady, VP of Android at CARS, during a virtual briefing with media members before the meeting.
Gemini surfaces in two main ways in Android Auto Experience.
Gemini acts as a much stronger smart voice assistant. The driver (or passenger – Brady said he didn’t do voice matching to anyone who owns a phone running Android Auto Experience) texts Gemini, asks him to play music, and basically can do everything Google Assistant can already do. The difference is that thanks to Gemini’s natural language capabilities, users don’t have to be that robotic with their own commands.
Gemini can “remember” things like whether a contact prefers to receive text messages in a particular language, and can also handle user translations. And Google claims that Gemini can run one of the most commonly paradeed in-vehicle tech demos. Finding a good restaurant along the planned route. Of course, Brady said Gemini could minify Google’s listings and reviews to accommodate more specific requests (such as “Taco Places with Vegan Options”).
Another main way Gemini surfaces is what Google calls “Gemini Live.” This is an option that digital AI is essentially constantly listening and ready to have a full conversation no matter what. Brady said these conversations were about everything from spring break travel ideas to brainstorming recipes for the 10-year-old’s desired recipes to “Roman history.”
If it all sounds distracting, Brady said Google believes it isn’t. He argued that natural language capabilities allow Gemini to “reduce cognitive load” so that they can ask Android Auto to perform certain tasks in a fuss.
It’s a bold claim to make when auto companies are screaming for people to get back physical knobs and buttons away from the touchscreen.
There’s still a lot to be sorted out. For now, Gemini leverages Google’s cloud processing to run on both Android Auto and on cars with built-in Google. But Brady said Google could work with the automaker to “incorporate more calculations (Gemnini) at the edge.” This helps not only performance but also reliability.
Modern cars also generate a lot of data from onboard sensors, as well as some models, interior and external cameras. Brady said that “there is nothing to announce” about whether Gemini can take advantage of its multimodal data, and that “we’re talking a lot about it.”
“With more and more cameras in cars, we definitely think there will be some really interesting use cases in the future,” he said.
Android Auto’s Gemini and Google built-in are already appearing in all countries that have access to the company’s generative AI models, supporting over 40 languages.
Learn how to watch live streams and more from Google I/O.