Gemini is rolling out new features and improvements, including enhanced performance, personalization, and expanded app integrations.
The 2.0 Flash Thinking Experimental model now offers a longer context window, allowing users to tackle more complex problems. This upgrade also supports file upload, making Gemini’s reasoning capabilities stronger.
Deep Research has been upgraded with the new 2.0 Flash Thinking Experimental model, providing higher-quality reports and a real-time look into Gemini’s thought process. The feature is now available globally in 45+ languages for free, with expanded access for Gemini Advanced users.
Gemini’s personalization feature has also been enhanced, connecting with Google apps and services to deliver more tailored responses. This allows users to ask questions like “What restaurants have you searched for recently?” or “Where did I last travel?”
Connected apps are now more powerful, with Gemini integrating with Google Calendar, Notes, Tasks, Photos, and upcoming Google Photos. Users can access these features in a single prompt, asking Gemini to find them a recipe, create a shopping list, or recall information from recent trips.
Gems, customizable AI experts, are also available for everyone at no cost. Users can create their own Gems using pre-made templates or write instructions to create a custom expert on any topic.
These improvements are now available in the Gemini app, with users able to try them out at gemini.google.com.
Source: https://blog.google/products/gemini/new-gemini-app-features-march-2025