Google is quietly reshaping how users interact with its Gemini app ahead of the highly anticipated I/O 2026 event, set for May 19–20. Among the latest updates, a new “Thinking level” option is now rolling out to a limited group of users, allowing them to toggle between “Standard” and “Extended” modes when using the Fast (Gemini 3 Flash) or Gemini 3.1 Pro models. This feature mirrors the adjustable reasoning levels already available in Google AI Studio, where users can select Low, Medium, or High thinking modes. The addition reflects Google’s ongoing effort to give users finer control over computational depth, balancing speed and complexity depending on the task at hand.
According to verified sources, the “Thinking level” menu appears only when selecting specific models, and its rollout is currently limited. This update comes as Google continues to position Gemini not just as a conversational assistant, but as a versatile digital agent capable of integrating with a growing ecosystem of third-party applications. The goal is to streamline workflows across creative, shopping, and dining experiences, making AI assistance more seamless and context-aware.
The new Gemini app features are part of a broader push to deepen third-party integrations, with support already confirmed for apps like GitHub, OpenStax, Spotify, and WhatsApp. Support documents further indicate that Canva, Instacart, and OpenTable are in development, though these integrations have not yet been rolled out to the public. Once available, these connections will allow users to perform tasks such as generating and editing designs in Canva, adding groceries directly to an Instacart cart, or booking and managing restaurant reservations through OpenTable—all via natural language prompts.
For creatives, the Canva integration promises to bridge the gap between idea generation, and execution. Users will be able to ask Gemini to create new designs, manage project files, and even respond to feedback within Canva. Example prompts range from generating a rustic wedding invitation to organizing marketing assets or replying to design comments. Meanwhile, the Instacart integration could transform grocery shopping into a fully automated process, with Gemini capable of identifying nearby stores, checking product availability, and adding items to a cart based on user instructions.
Similarly, the OpenTable integration leverages Google’s Reserve with Google service to check restaurant availability, book tables, and manage reservations. Users can ask Gemini to find a suitable restaurant, book a table, and even sync the reservation with their Google Calendar. The system also supports cancellations and updates, providing a hands-off way to manage dining plans.
What the ‘Extended’ Thinking Level Means for Users
The introduction of “Extended” thinking level is designed to address scenarios where users require more in-depth analysis or multi-step reasoning. For instance, a user researching a complex topic or planning a multi-variable project might benefit from the additional computational effort provided by the Extended mode. Conversely, those prioritizing speed—such as drafting quick emails or summarizing articles—can opt for the Standard level. This flexibility aligns with Google’s broader strategy to make AI tools more adaptable to individual needs and workflows.
While the rollout is limited, the addition of thinking levels and new integrations signals Google’s ambition to turn Gemini into a central hub for productivity and daily tasks. The company has previously announced plans to expand Gemini’s capabilities at I/O 2026, and these updates suggest further innovations are on the horizon. For now, users interested in testing the new features should keep an eye on their Gemini app for notifications or updates.
How to Prepare for the Full Rollout
As Google continues to refine and expand Gemini’s functionality, users can expect more integrations and features in the coming months. The company has confirmed that I/O 2026 will highlight additional AI breakthroughs, including updates to Gemini and other products. Until then, those eager to explore the latest capabilities can experiment with the current thinking levels and existing third-party integrations, keeping in mind that availability may vary by region and user group.
For developers and power users, the introduction of adjustable reasoning levels and deeper app integrations opens new avenues for automation and workflow optimization. As these features become more widely available, they may redefine how individuals and businesses leverage AI for everyday tasks—from creative projects to personal errands.
Google has not yet announced a specific timeline for the full rollout of Canva, Instacart, and OpenTable integrations, but the company’s track record suggests these updates will follow in the near future. In the meantime, users are encouraged to stay updated through official Google channels, including the Gemini app release notes and the I/O 2026 event page.
As Gemini evolves, one thing is clear: the app is becoming more than just a chatbot. It’s a digital assistant that understands context, connects with the tools we use daily, and adapts to how we work. The next steps will determine just how seamlessly it can integrate into our lives.
Have you tried the new Gemini features? Share your experience in the comments or let us know what integrations you’d like to see next.
