OpenAI has debuted a new workspace interface for ChatGPT called Canvas. The AI giant announced the new ChatGPT workspace on its official blog and immediately made it available to ChatGPT Plus and Team users. Enterprise and Edu users will be able to access Canvas within the next week.
Canvas is a virtual interface space for creating and coding projects, and users can discuss specific parts of their projects using ChatGPT. According to the blog, a separate window opens next to the main chat space where users can write and type code into this new “canvas,” highlight sections to focus on the model, and use the “copy editor and code You can edit it like a reviewer.
Canvases can also be opened manually by typing “use Canvas” at the prompt. According to the blog post, Canvas can also open automatically when it “detects a potentially useful scenario.” There are also several shortcuts you can use to create and code projects. For writing projects, users can ask ChatGPT to suggest edits, adjust length, or change the reading level of a block of text from graduate level to kindergarten level. You can also add “emoticons related to emphasis or color.”
Code authors can have ChatGPT review their code and add inline suggestions for improvement. You can also mark up your work with logs and comments to aid debugging and make your code easier to understand. You can also fix bugs and port your coding to another language like JavaScript, TypeScript, Python, Java, C++, or PHP in canvas mode.
OpenAI’s Canvas feature allows ChatGPT to work with other AI assistants, such as Anthropic’s Artifacts and the coding-focused AI model Cursor, which have separate workspaces to focus on specific areas of a project.
Updated October 4, 12:55 PM ET: This story was edited after publication to include more context about the Canvas feature code and text features.
