Google Gemini Expands Capabilities: New ‘Thinking’ Controls and Major Third-Party Integrations

Google is moving rapidly to transform Gemini from a conversational chatbot into a proactive digital agent. Recent updates to the Gemini app signal a dual-track strategy: increasing the depth of its reasoning capabilities and expanding its reach through a growing ecosystem of third-party integrations. From fine-tuned “thinking” levels to seamless connections with apps like Canva and Instacart, the evolution of Gemini is aimed at making AI a central orchestrator of daily productivity and lifestyle tasks.

Fine-Tuning Intelligence: The Arrival of ‘Thinking’ Levels

One of the most significant technical shifts appearing in the Gemini app is the introduction of a “Thinking level” option. This feature allows users to control the computational depth the model applies to a specific query, providing a more nuanced way to balance speed, and reasoning.

Currently, this menu is appearing in a limited rollout and is specifically available when users select the Prompt (Gemini 3 Flash) or Gemini 3.1 Pro models. When these models are selected, a new menu appears at the bottom of the model picker, offering two distinct modes:

• Standard: Optimized for quick responses and general tasks.
• Extended: Designed for more complex reasoning and deeper analysis.

This functionality mirrors the existing controls found in Google AI Studio, which offers Low, Medium, and High thinking levels. By bringing these controls to the consumer-facing Gemini app, Google is giving users more agency over how much “mental effort” the AI exerts on their behalf.

Building an Agentic Ecosystem: Third-Party Integrations

Beyond improving how the model thinks, Google is focused on what the model can do. The Gemini app is expanding its ability to interact with the external world through third-party app integrations. While users can already utilize integrations with @GitHub, @OpenStax, @Spotify, and @WhatsApp, a new wave of service-oriented connections is on the horizon.

Support documentation reveals that Google is preparing to roll out deep integrations with Canva, Instacart, and OpenTable, which will allow Gemini to act as a bridge between user intent and service execution.

Creative Workflow with Canva

The upcoming Canva integration is designed to streamline the creative process by connecting Gemini’s generative power with Canva’s design tools. Users will be able to manage entire design lifecycles through simple prompts. Capabilities include:

• Design Generation: Creating new designs based on text prompts (e.g., “Create a rustic wedding invitation with a dusty rose palette”).
• Asset Management: Accessing existing designs, templates, and project files, or moving assets between different folders.
• Collaborative Editing: Listing and replying to comments on specific designs.
• Cross-Platform Workflow: Generating an image within Gemini and immediately using it to build a business card or other design in Canva.

Grocery Management with Instacart

The Instacart integration aims to turn Gemini into a personal shopping assistant. By leveraging the user’s default saved address to find nearby stores, Gemini can facilitate seamless grocery and household item shopping. Users can:

• Automate Shopping Lists: Ask Gemini to “add all the ingredients” from a specific recipe directly to an Instacart cart.
• Check Availability: Query specific products or find which local stores carry certain items, such as vegan cheese.
• Specific Product Selection: Specify exact versions of products or preferred stores during the add-to-cart process.

Dining and Reservations with OpenTable

Through a connection with Reserve with Google, Gemini will be able to manage dining arrangements via OpenTable. This integration moves Gemini into the realm of real-world logistics, allowing users to:

• Search and Book: Find restaurants based on cuisine, group size, or time availability and book them instantly.
• Manage Reservations: Check existing bookings, change times, or cancel reservations through natural language commands.
• Calendar Integration: Automatically add confirmed reservations to a Google Calendar and set reminders.

Key Takeaways

• Enhanced Control: The new “Standard” and “Extended” thinking levels allow users to choose between speed and deep reasoning in Gemini 3 Flash and 3.1 Pro.
• Action-Oriented AI: Upcoming integrations with Canva, Instacart, and OpenTable shift Gemini from a text generator to an “agent” capable of executing real-world tasks.
• Workflow Integration: The focus is on reducing friction, whether that is moving design assets in Canva or managing dinner plans via OpenTable.

Frequently Asked Questions

Which Gemini models support the new “Thinking level” option?

The “Thinking level” menu is currently appearing for users selecting the Fast (Gemini 3 Flash) and Gemini 3.1 Pro models.

Can I use Gemini to manage my Canva folders?

Yes, once the integration rolls out, you will be able to ask Gemini to perform tasks like creating new folders or moving presentations to specific folders within Canva.

How does Gemini handle OpenTable bookings?

Gemini uses Reserve with Google to check real-time availability and complete bookings, changes, or cancellations through the OpenTable platform.