Google’s Gemini-Powered AI Mode Can Now Find and Book Restaurant Tables in India

Google is rolling out new agentic capabilities in its AI Mode, starting with restaurant reservations in India, as the company pushes its Gemini AI beyond search and into task execution.

The feature allows users to ask for restaurant bookings with multiple constraints such as location, cuisine, party size, date, and time. Instead of returning a list of links, AI Mode processes these inputs and surfaces restaurants with real-time table availability.

AI Mode Moves From Search to Action

With this update, Gemini AI is no longer limited to answering queries or suggesting options. It now completes part of the task. Users can input a detailed request, such as planning a dinner with specific preferences, and AI Mode will scan multiple platforms to identify suitable options. It then presents a curated list of restaurants with available reservation slots.

Once a user selects an option, the system redirects them to the booking page to complete the reservation.

Aggregating Across Platforms

The system works by pulling data from multiple sources rather than relying on a single platform.

In India, Google has partnered with services including Swiggy, Zomato, and EazyDiner to enable real-time discovery of restaurant availability. This allows AI Mode to compare options across platforms and present consolidated results.

Powered by Live Web and Google Stack

Under the hood, the feature uses live web browsing capabilities from Project Mariner, along with integrations within Google Search. It also draws on Google’s Knowledge Graph and Maps data to refine results based on location and relevance.

A Step Toward Agentic AI

The rollout reflects Google’s broader push toward “agentic” AI systems that can perform tasks rather than just assist with information. Restaurant booking is an early use case, but it signals how Gemini AI is being positioned to handle more complex, multi-step actions directly within Search.

For now, the feature is being rolled out in India, with a focus on dining use cases that involve real-time availability and multiple decision variables.

Read more: 10 Petabytes Stolen, Missiles Exposed? Hacker Claims Chilling Breach of China’s Top Secret Supercomputer