OpenAI makes ChatGPT Canvas accessible to everyone: How to use
OpenAI has announced that its digital editing platform, Canvas, is now available for all ChatGPT users. The announcement was made during a livestream event as part of the "12 Days of OpenAI" series. Previously limited to paying subscribers of ChatGPT Plus, Teams, Edu and Enterprise plans, Canvas is now available to all web users in GPT-4o.
Canvas: A new approach to digital editing
Canvas reimagines the classic ChatGPT interface into a more intuitive design. It displays a conversation on the left-hand sidebar and a workspace for user projects on the right side of the chat session screen. This way, users can request changes from ChatGPT, which will be automatically reflected in the content area, without having to generate a completely new text response.
OpenAI integrates Canvas into GPT-4o
OpenAI has integrated Canvas into GPT-4o, making it a native feature of the model. This means users won't have to switch between models using the model picker. The updated version of Canvas is designed to automatically open for certain prompts or pasted text, further enhancing user convenience.
Canvas: A tool for code execution and debugging
The latest version of Canvas can run Python code, handle more text pasting, and be launched in custom GPTs. It can even let you paste Python code into ChatGPT, which could automatically trigger the opening of Canvas. Apart from these, OpenAI also demonstrated that Canvas can generate and preview graphics from code alone. This way, developers or analysts can tweak formulas or data before finalizing a chart.
Canvas's role in custom GPTs
For those creating custom GPTs, Canvas will be a default feature. However, users can decide when and if Canvas opens for prompts on their custom assistant. For existing custom GPTs, OpenAI hasn't made Canvas a default feature to not disrupt their current functionality. Users can add Canvas as a feature to their GPTs through the settings of the custom GPT.