Google Search Revolution: Gemini 3.5 Flash Powering New Agentic AI Features
The New Era of Search: Gemini 3.5 Flash Integration
Google has officially transitioned Google Search to be powered by Gemini 3.5 Flash, marking a significant leap in the efficiency and capability of its AI-driven experiences. According to Google, Gemini 3.5 Flash is designed to provide “sustained frontier performance,” specifically optimized for complex agentic tasks and advanced coding. This model is now the global default for users operating in AI Mode, ensuring faster, more intelligent, and more context-aware search results for millions of users worldwide.
The Rise of Information Agents
One of the most ambitious updates is the introduction of Information Agents. Unlike traditional search queries that provide a one-time answer, these agents operate in the background 24/7. They are designed to monitor the web—including blogs, news sites, social media, and real-time data streams—to track ongoing projects or specific interests for the user.
For example, a user hunting for an apartment can provide a comprehensive list of requirements, and the agent will continuously scan listings, notifying the user the moment a matching property appears. Similarly, sports fans can set alerts for specific athlete collaborations or product drops. These features are slated to launch first for Google AI Pro and Ultra subscribers this summer.
Agentic Booking and Personal Intelligence
Google is also broadening its Agentic Booking capabilities. Users can now delegate real-world logistics to Google, such as calling businesses to book home repairs, beauty appointments, or pet care services. This rollout is expected to hit the U.S. market throughout the summer.
Furthermore, Personal Intelligence is expanding its reach. Originally launched in Gemini and AI Mode earlier this year, this feature is now being rolled out to nearly 200 countries and territories across 98 languages. Notably, this expansion does not require a paid subscription, making a more personalized search experience accessible to a global audience.
Agentic Coding and Generative UI
In a move that blurs the line between search and software development, Google has introduced Agentic Coding powered by Google Antigravity and Gemini 3.5 Flash. This allows users to customize how search results and widgets behave and look through two primary methods:
- Generative UI: Search can now design custom layouts in real-time, assembling interactive visuals, tables, graphs, and simulations on the fly. This will be available for free to all users this summer.
- Custom Coding Experiences: For complex, ongoing tasks like wedding planning or managing a home relocation, users can build “mini-apps” or custom dashboards. For instance, a user could create a custom fitness tracker that integrates real-time weather, local maps, and reviews. This will debut for AI Pro and Ultra subscribers in the U.S.
Universal Cart: Reimagining E-commerce
Shopping is getting a massive overhaul with the Universal Cart. This feature allows users to add products from various retailers (such as Nike, Sephora, Target, and Walmart) into a single unified cart. Users can check out via Google Pay or transition to the individual merchant’s site.
The Universal Cart is powered by Gemini to provide intelligent insights, such as flagging product incompatibilities (e.g., when building a custom PC) and tracking price history to alert users of drops. This is rolling out in the U.S. this summer across Search and the Gemini app, with YouTube and Gmail integration to follow.
Secure Transactions with AP2
To support these autonomous actions, Google has developed the Agent Payments Protocol (AP2). This protocol ensures that agentic payments are secure and transparent, providing a verifiable digital paper trail between the user, the merchant, and the payment processor. AP2 includes strict guardrails to ensure agents only act on the user’s behalf, starting with its integration into Gemini Spark.