Google Brings Gemini’s Canvas in AI Mode to US Search

Google is rolling out Gemini’s Canvas inside AI Mode in Search to all users in the United States, turning the search box into an AI workbench for planning trips, prototyping apps and organizing complex projects. The feature, available in English, promotes Canvas from a niche Gemini playground into a mainstream tool that appears directly above traditional results when people opt into AI Mode.

The expansion means millions of users who may never open a dedicated AI app will now encounter a persistent, multimodal workspace woven into everyday search flows, according to coverage from TechCrunch and other outlets.

What Canvas in AI Mode actually does

Canvas is an interactive workspace that lives inside AI Mode, Google’s conversational layer on top of Search that is powered by its Gemini models. Once in AI Mode, users can tap a Canvas icon or "+" button to open a side panel that stays active as they keep searching, letting them refine plans, drag in new information and iterate on code or content in one continuous environment, as explained by search-focused site The Search Signal.

Unlike a single-shot answer box, Canvas is designed for multi-step work. Walkthroughs from PCWorld and other reviewers show users asking AI Mode to generate a simple budgeting or workout app, then using Canvas to inspect and edit the generated code, preview basic interfaces and iterate on features without leaving the search results page.

Canvas also supports non-technical tasks like research organization, trip planning and document drafting. A November Google AI update highlighted Canvas as a way to “describe the kind of trip you want and what recommendations you need,” then turn those suggestions into structured itineraries within Search, according to a company blog post on recent AI features in Search and Gemini’s tools for planning and shopping (Google).

Why this matters for mainstream AI use

By putting Canvas directly into Search, Google is effectively collapsing the gap between asking a question and building something with the answer. Analysts at the Digital Watch Observatory note that bringing Canvas into AI Mode gives users research tools, coding assistance and document creation “directly within search,” reframing Search as an execution environment rather than just a list of links (Digital Watch Observatory).

That shift could have real consequences for who gets to use advanced AI in practice. Instead of requiring people to seek out a separate IDE, low-code platform or standalone chatbot, AI Mode now offers a no-setup path to generate spreadsheets, prototype web pages or sketch business plans from natural-language prompts. Guides from independent publishers like Just Think AI argue that this integration lowers the bar for non-technical users to move from “searching for how” to “actually doing” complex tasks such as app prototyping and multi-day project planning inside a single Google tab (Just Think AI).

The move also tightens competition among AI assistants. Microsoft has been pushing Copilot and its own canvas-like experiences inside Bing and Office, while startups like Notion and Replit offer embedded AI workspaces in productivity and coding tools. By routing creation workflows through Search itself and leaning on Gemini’s multimodal reasoning, Google is betting that the default place people go to ask questions can also become the default place they build.

What changes for users and developers

For everyday users in the US, the immediate change is that entering AI Mode in Search now exposes a visible Canvas option without requiring a separate Gemini app or Workspace subscription, as reported by TechCrunch. That makes it easier to save, revisit and expand on AI-generated work over multiple queries, instead of copying text into documents or IDEs.

For developers and power users, Canvas in AI Mode offers a lightweight on-ramp for rapid prototyping and code exploration tied directly to web results. Google’s own AI updates have framed Canvas as part of a broader push toward “generative UI” — interfaces that change dynamically based on what the model is doing — and as a companion to deeper tools like Gemini Agent and Deep Research for handling multi-step tasks (PPC Land).

Unanswered questions remain about how far Google will take this approach, including whether Canvas workspaces will become more shareable or hook more deeply into services like Sheets, Docs and Android. But by moving Canvas into the line of sight of mainstream search users, Google is signaling that AI in Search is no longer just about better answers — it is about turning the results page into a place where work actually gets done.