San Francisco, California – After a series of developer announcements on Tuesday, Google has unveiled several updates to its Gemini app for end users. The updates are centered around an enhanced version of the 2.0 Flash Thinking Experimental, which is currently in preview.
One notable update is the introduction of more Gemini Advanced features for free users. Google has made document upload and the generation of images of people available to all users. In addition, free users over the age of 18 can now use premade Gems or create their own custom versions of Gemini for specific tasks.
These Gems can be accessed on mobile devices, but the ability to create new ones is limited to the web. The Gems manager allows users to write instructions, upload files for reference, and assign a name to each customized Gem. Google is also making Deep Research available to all free Gemini users in over 45 languages.
The new version of Gemini 2.0 Flash Thinking Experimental promises better efficiency and speed. With an improved reasoning model, Gemini can now break down prompts into a series of steps to enhance its reasoning capabilities and provide more accurate responses. Advanced users can now upload files and access a 1 million token context window with the model.
Moreover, the 2.0 Flash Thinking Experimental Model is now compatible with a wide range of Gemini apps, allowing users to seamlessly integrate services like Gmail, Google Calendar, Drive, Messages, YouTube, and more. This integration enables users to make complex requests that involve multiple apps, with Gemini reasoning over the entire request to provide comprehensive responses.
Google is planning to introduce a Google Photos app in the coming weeks to facilitate “Ask Photos” functionality. The app will allow users to ask Gemini to analyze photos and provide relevant information, such as creating a travel itinerary based on recent trips or recalling specific details.
Additionally, Google is using the 2.0 Flash Thinking Experimental to power a new “Personalization (experimental)” model in Gemini. This model will leverage data from Google apps and services to deliver responses tailored to individual user needs. For example, Gemini can provide personalized restaurant recommendations based on recent searches or offer travel advice based on previous destinations searched.
The Personalization model is gradually rolling out to web users before expanding to mobile users in the future. With this new update, Gemini aims to provide more personalized insights and responses by drawing from a user’s activities and preferences across various Google services.