OpenAI launched Canvas, a new way to collaborate with ChatGPT
OpenAI has launched "canvas," a new beta workspace feature for ChatGPT that allows users to collaborate on writing and coding projects through a dedicated interface with specialized tools.
OpenAI's most recent product is a workspace it calls 'canvas', which is independent of the chat window and lets users collaborate with ChatGPT on writing and coding projects. OpenAI's canvas offers similar functionalities as Anthropic's Artifacts, which is expected given the growing consensus that although chatbots like Claude and ChatGPT may be able to provide acceptable sketches for coding and writing projects, prompts alone aren't sufficient to work on iterating and improving on these starting points.
Currently an early beta for Plus and Team subscribers, the product will be available for Edu and Enterprise users soon. OpenAI has shared its plans to give free tier users access to ChatGPT with canvas once it is out of beta. Eligible users must choose the "GPT-4o with canvas" model manually from the model picker, which suggests that OpenAI eventually expects to make it so ChatGPT automatically understands when to launch a canvas without the additional step of loading a specific models, in a similar way as Claude automatically launches the Artifacts panel.
The canvas interface provides separate sets of tools to collaborate on writing or coding projects. In writing projects, canvas provides shortcuts for ChatGPT-generated inline suggestions and edits, adjusting the length of the text, adjust the complexity (from Kindergarten to graduate school level), doing a final check, and adding emoji. Coding-specific canvas shortcuts enable users to ask ChatGPT to review code, add logs and comments, fix bugs, and port existing code to a language.
