Google is rolling out new updates to the Gemini app ahead of Google I/O 2026. The updates include a new setting for how the AI processes prompts, as well as major new connections to popular third-party services that most users have deeply integrated into their workflows.
Gemini App Now Lets You Adjust the Thinking Level
Spotted first by 9to5Google, Google is rolling out a new Thinking Level menu at the bottom of the model picker sheet in the Gemini app. This menu only shows up when users select the Fast model powered by Gemini 3 Flash or Gemini 3.1 Pro.
Users can choose between Standard and Extended thinking. This allows them to give Gemini more time to process complex requests, similar to the Low, Medium and High Thinking options found in the Google AI Studio. This feature is gradually rolling out to all users.
That said, we also spotted an auto-switching feature earlier today, where Gemini used the "Thinking" model instead of Pro because the demand for the latter at the point in time was much higher. That said, Google is soon extending support for more apps.
One of the Gemini features includes Gemini extensions and Personal Intelligence, where Gemini can get user info and do tasks on apps like GitHub, OpenStax, WhatsApp and more, but according to the latest information, the AI app is being prepped to support three more major integrations. These include Canva, Instacart and OpenTable.
Users connecting Canva to Gemini will be able to streamline their design workflows and generate ideas using Gemini. Instacart users can check restaurant availability and complete bookings through OpenTable with prompts like "What stores near me carry vegan cheese" or "I want to make this. Can you add all the ingredients to my Instacart?"
With Google I/O 2026 set for May 19, 2026, we should see Google reveal more about these new integrations on stage. Besides, the stage is also expected to be filled with tons of new Gemini announcements and features. Stay tuned to Beebom Gadgets to learn more about all the upcoming features.
