What you need to know
- Google’s massive announcement involves bolstering the usefulness and complex query solving capabilities of the Gemini app.
- The post highlights a 2.0 Flash Thinking Experimental model upgrade, which aims to enhance Gemini, as well as bring new features.
- The Gemini app picks up a new “personalization” experimental feature, which lets the AI scour your Google app/service history for “tailored” responses.
- Google is also letting users try out its latest Deep Research updates in Gemini, but only for a limited time a month.
Today (Mar. 13), Google highlighted a host of Gemini app updates and features users can soon expect.
In a blog post, Google announced a series of improvements en route for Gemini, beginning with the 2.0 Flash Thinking Experimental model. According to the post, this upgrade for the model gives Gemini the ability to accept “new features,” such as file uploads. More importantly, Google states the upgraded model gives Gemini more capabilities to break down prompts into “a series of steps” for better responses.
The 2.0 Flash Thinking Experimental model also gives Gemini Advanced users a 1M context token window to understand/compute complex problems.
Gemini App Updates
Image 1 of 2
Those who utilize the Gemini app regularly are in for a more personalized treat as Google rolls out a major update. The company says, beginning today (Mar 13), users will notice an experimental feature called “personalization.” This is powered by the upgraded 2.0 Flash Thinking Experimental model to connect your Google apps and services with Gemini.
The company states this is done to facilitate a “tailored” experience that should fit your needs, should you desire. The test will initially begin by leveraging your behavior in Search. As an example, Google states users looking for restaurant recommendations will see Gemini dive into your Search history for anything food-related.
Google states users can enable the personalization test in Gemini by selecting its option via the “Model” drop-down menu. The post adds that users remain in control of Gemini and can disable this feature whenever they like.
Speaking of leveraging connecting apps, Google says it’s bringing more services to Gemini, like Calendar, Notes, Tasks, and Photos via 2.0 Flash Thinking Experimental. Due to this, Gemini is said to have better capabilities to process complex queries that involve multiple Google apps. As an example, a user could ask Gemini to discover a recipe on YouTube, add the necessary ingredients to a shopping list, and find the nearest store.
Google then teased additional integration between Gemini and the Photos app, so users can ask for help/questions “based on your photos.”
Lastly, Google announces that Gemini’s Gems — which lets users create their own personalized AI — are rolling out for “everyone at no cost” in the app.
Deep Research & Gemini
Wrapping up Gemini’s haul of updates is another for Deep Research; a model originally introduced as a way for users to explore complex topics with easy to read reports. Today, Google states Deep Research will benefit from the 2.0 Flash Thinking Experimental model for “enhanced” functions.
Google says this will concern “all stages:” planning, searching, and reasoning, as well as analyzing and reporting.
With this upgrade, users will discover deeper insights into how Gemini thinks while it’s attempting to fulfill your query. Users will see, in real-time, Gemini’s web searches and how it compiles the information for your viewing pleasure once it’s done.
To celebrate the update, Google announces Deep Research is coming to all users globally within the Model drop-down menu in Gemini. However, Google states users can “try Deep Research a few times a month at no cost, and Gemini Advanced users get expanded access to Deep Research to save even more time on their most complex projects.”