Google is rolling out a suite of upgrades to its Gemini AI, bringing enhanced research capabilities, a longer context window, and deeper integration with Google services. These updates aim to make Gemini a more powerful and personalized assistant, whether users are tackling complex research projects or looking for smarter daily assistance.
Smarter Reasoning with 2.0 Flash Thinking Experimental
The new 2.0 Flash Thinking Experimental model enhances Gemini’s ability to break down complex queries into structured steps, improving both reasoning and response quality. This update brings file upload support, faster processing, and a 1M token context window for Gemini Advanced users. The extended context allows Gemini to analyze larger datasets and generate more insightful responses for intricate problem-solving tasks.
Deep Research Expands to More Users
Google’s Deep Research tool, first introduced in December, has been upgraded with the new 2.0 Flash Thinking Experimental model. This AI-powered research assistant searches, analyzes, and synthesizes information from across the web, delivering detailed multi-page reports in minutes.
One key improvement is real-time transparency—users can now see Gemini’s thought process while it browses the web, offering a clearer look at how it approaches research tasks. Starting today, Deep Research is rolling out globally in 45+ languages. Standard Gemini users can access it a few times per month at no cost, while Gemini Advanced users get expanded access for more in-depth projects.
Personalized Responses with Search Integration
Google is introducing personalization to Gemini, making interactions more tailored to individual users. By connecting with Google services—starting with Search—Gemini can refine its recommendations based on previous searches.
For instance, users asking for restaurant suggestions will receive options influenced by their recent food-related searches. Similarly, those seeking travel advice will see responses shaped by past destination queries. This feature, labeled Personalization (experimental), can be enabled or disabled at any time, giving users full control over their search history and data preferences.
Deeper Integration with Google Apps
Gemini is expanding its connected app support to include Calendar, Notes, Tasks, and Photos, with English availability starting today. This update allows Gemini to handle more complex, multi-step requests that involve multiple services.
For example, a user can ask:
“Find an easy cookie recipe on YouTube, add the ingredients to my shopping list, and locate grocery stores still open nearby.”
Additionally, Google Photos integration is coming soon, enabling Gemini to assist with image-based requests. Users will be able to generate travel itineraries from past vacation photos or quickly retrieve important information, such as the expiration date on a driver’s license.
Gems Now Available for Everyone
Google is making Gems, its AI-powered custom assistants, available at no cost to all Gemini users. Gems allow users to create specialized AI helpers for different tasks, such as language translation, meal planning, or math tutoring.
Users can either start with pre-made Gems or create their own by setting instructions and uploading relevant files. Once configured, these AI assistants can be accessed instantly within the Gemini app.
Try the New Gemini Features Now
Google continues to refine and expand Gemini’s capabilities, making AI-driven assistance more powerful and accessible. The latest updates are rolling out now, with more features on the way. Users can start exploring these enhancements at gemini.google.com.
