Google is expanding Gemini Canvas inside AI Mode in Search to all users in the United States, giving more people a built in workspace for planning, writing, and coding.
Quick Summary – TLDR:
- Canvas in AI Mode is now available to all US users in English.
- It adds a dedicated workspace to draft documents, organize projects, and build interactive tools.
- Users can create prototypes, view the code, and refine results through chat with Gemini.
- The move puts Google’s Canvas in more direct competition with OpenAI and Anthropic.
What Happened?
Google has rolled out Canvas in AI Mode to everyone in the US, expanding the feature beyond earlier limited tests inside Google Labs. Canvas now supports creative writing and coding, alongside planning and research style tasks, all from within Google Search.
Canvas in AI Mode lets you build plans and organize information over multiple sessions in a dynamic side panel that updates as you go. Now we’re making this tool available to everyone in the U.S. in English, and we’re adding support for creative writing and coding tasks, so you…
— Google (@Google) March 4, 2026
Canvas Turns Search Into a Workspace
Canvas is meant to be more than a one off answer box. Instead, it acts like a persistent space where you can organize ideas, keep building on a project over time, and pull in fresh information while you work.
Google says Canvas can help with:
- Planning projects and organizing research
- Drafting documents inside Search
- Creating custom tools and interactive dashboards
- Refining creative writing drafts with feedback
One example Google highlighted from early testers was a dashboard to track academic scholarships, pulling together requirements, deadlines, and dollar amounts in one place.
How to Use Canvas in AI Mode?
The workflow is simple and very “inside Search.”
To start:
- Open AI Mode in Google Search.
- Select the plus button inside the chat window.
- Choose Canvas from the menu.
- Describe what you want to create.
Canvas opens in a side panel on the right, where it can combine information from the web and Google’s Knowledge Graph. If you are building a prototype or app, you can test how it works, toggle to view the underlying code, then refine the result with conversational follow ups until it matches what you want.
From Study Guides to Apps and Games
Google is positioning Canvas as useful for both everyday tasks and deeper work. Earlier examples included building a study guide by uploading class notes and other sources. It can also transform a research report into formats like a web page, a quiz, or an audio overview, which overlaps with the kind of workflow people use in tools like NotebookLM.
On the creation side, Canvas can generate code for an idea and turn it into something shareable, like an app or game, then let you iterate quickly by chatting with Gemini.
Why Google Is Pushing This Through Search?
Canvas already exists inside the Gemini app, where Google AI Pro and Google AI Ultra subscribers can access the latest model, Gemini 3, plus a larger 1 million token context window for complex projects.
The bigger story is distribution. By placing Canvas inside Search, Google is giving the feature a much wider audience, including people who may not have used Gemini directly. Search remains Google’s strongest advantage, because it can introduce new AI features at massive scale.
The Competitive Angle: OpenAI and Anthropic
Canvas is also part of a growing category of AI workspaces. Google’s approach competes with similar tools from OpenAI and Anthropic.
One difference is how the feature is triggered. ChatGPT Canvas can open automatically based on what you ask. Google’s Canvas and Anthropic’s Claude generally require a more direct action, like choosing a tool option, before you land in that workspace style experience.
SQ Magazine Takeaway
I think this is Google doing what it does best: using Search as a delivery system. A lot of AI tools are impressive, but they live behind separate apps and subscriptions. When a feature like Canvas sits inside the place people already use every day, it has a real shot at becoming normal. If Google keeps improving the writing and coding experience here, it could quietly become the default way many people plan projects and build simple tools, without ever thinking of it as “using AI.”
