Google AI Mode Gets UI Redesign as Restaurant Booking Expands Globally

Google has rolled out a redesigned interface for AI Mode on mobile devices and expanded its agentic restaurant booking capability to 8 new countries, moving beyond its initial US-only launch.

UI Redesign Details

The prompt box interface is being replaced with a bottom sheet layout on both Android and iOS, aligning the mobile experience more closely with web. The new design features:

• Large buttons for Gallery and Camera at the top
• "Tools" section with "Create images" as the currently available mobile option
• "Gemini 3 models" switcher allowing selection between Auto or Pro versions
• Elimination of the previous pop-up menu approach

The redesign mirrors recent changes to the main Gemini app interface and is rolling out incrementally to the stable channel across both platforms.

Restaurant Booking Goes International

Google's agentic booking feature—which autonomously books restaurant reservations by searching multiple reservation platforms and local partnership websites—is now available in 8 new markets:

• Australia
• Canada
• Hong Kong
• India
• New Zealand
• Singapore
• South Africa
• United Kingdom

The capability requires no Labs opt-in and accepts natural language queries specifying time, location, cuisine type, party size, or dining atmosphere preferences. Example queries include "Find a table for two at a dog-friendly Italian restaurant in Shoreditch for Saturday at 7 p.m." and "Find me a sushi restaurant nearby that has a table for four that also serves vegan tempura."

The feature integrates with local reservation platforms through partnerships established in each market to surface and book available tables directly within AI Mode.

What This Means

Google is steadily positioning AI Mode as a practical agent for real-world transactions beyond conversation. The redesigned interface reduces friction for accessing key functions (camera, gallery, image generation, model selection), while international expansion of booking signals confidence in the feature's reliability and infrastructure scalability. The lack of Labs requirement for booking indicates Google considers this capability ready for general availability rather than experimental.