OpenAI’s Canvas, a collaborative interface providing visual previews of AI responses, is now accessible to all ChatGPT users, including those on the free tier. Previously a beta feature for Plus and Teams subscribers, Canvas is now fully integrated into the GPT-4 model, allowing it to run natively within ChatGPT.
This integration eliminates the need to specifically select Canvas from the model list. Furthermore, users can now directly input text and code, including Python, into the Canvas interface for analysis and improvement suggestions by ChatGPT. This streamlined process replaces the previous method of uploading through the chat interface and then loading it into Canvas.
ChatGPT
The enhanced Canvas also supports OpenAI’s custom GPTs, simplifying their design and development without requiring extensive programming expertise. A “Show Changes” feature highlights recent modifications to text or code, aiding in error identification and correction.
Canvas is accessible via the web and the Windows desktop application. OpenAI plans to continue refining and expanding Canvas functionalities in the near future.
This announcement coincides with OpenAI’s “12 Days of OpenAI” campaign. Recent releases include the Sora video generation system, capable of producing 20-second 1080p clips, and the full version of the o1 reasoning model family, available through a new $200-per-month Pro subscription. The broader availability of Canvas underscores OpenAI’s commitment to enhancing user experience and accessibility across its product offerings.
