Gemini in Google Maps: how AI planned a full-day city trip and delivered

Gemini in Google Maps is no longer an experiment — it has become genuinely useful. A Verge journalist tested the AI assistant in day-planning mode and reached a conclusion he didn't expect: the tool works. Gemini appeared in Maps relatively recently — unlike Gmail, where it has been present for over a year, often annoying users.

Maps is different. Here the AI is embedded in a context where it is truly needed: when you don't know where to go and want a specific route, not general advice. For the test, the journalist set an intentionally specific task: find playgrounds near the new light rail line, restaurants with a transport theme suitable for children, and several other locations.

After roughly an hour of working with the assistant — clarifying requests, additional filters, route revisions — a dozen places ended up in his bookmarks. Some recommendations turned out to be predictable: popular locations that regular search also returns. But several spots the journalist wouldn't have found on his own — Gemini handled non-trivial criteria and suggested options that normal ranking algorithms don't push to the first page.

That's the difference between a query "café nearby" and a conversation with an AI that understands context. Technically, Gemini in Maps works as a conversational interface on top of the familiar mapping engine. Users formulate a request in free form — the AI interprets it, suggests options, clarifies details.

You can ask follow-up questions, add constraints, change priorities. The result is saved as a route or list of places that can be manually edited. The approach has clear limitations.

Gemini relies on Maps data rather than real-time accuracy: establishments can be closed, hours outdated, ratings inaccurate. AI doesn't replace verification — it speeds up creating an initial list. It's important to understand this to avoid disappointment.

Nevertheless, the overall conclusion is unambiguous: Gemini in Maps does exactly what a good AI assistant should do. It reduces time on routine searches and adds value where simple algorithms fall short. This isn't a revolution — it's a convenient tool that really works.

For Google, this is an important signal. The company is embedding Gemini everywhere, and not everywhere is this perceived positively. Maps is one of rare cases where integration is justified by the task itself: it is precisely here that contextual understanding of requests brings obvious practical benefit.