On the eve of Google IO, the Gemini app is receiving some notable updates. According to 9to5Google, a new “Thinking level” feature is being introduced for select users, allowing choices between Standard and Extended reasoning modes while using Gemini 3 Flash or Gemini 3.1 Pro. The Extended mode caters to more complex topics that require extra processing time, mirroring a feature Google AI Studio offers developers. Additionally, Gemini is broadening its support for third-party apps. While GitHub, OpenStax, Spotify, and WhatsApp are already integrated, 9to5Google discovered documentation indicating forthcoming support for Canva, Instacart, and OpenTable.
The Canva integration will enable users to create and modify designs via Gemini prompts. Instacart will allow users to add ingredients to a cart through a recipe link or list. OpenTable will facilitate restaurant searches, bookings, and reservation management directly in the chat, including passing a confirmed reservation to Google Calendar. These integrations have yet to roll out, but the timing suggests a connection to Google’s upcoming developer conference. This may signal Google’s shift of Gemini AI towards becoming a more autonomous AI rather than merely a chatbot.