The new rendering and model capabilities move Canvas from an editing panel toward a more complete place to build and inspect interactive work, sharpening the product overlap with Claude.
First-order effects
ChatGPT users can use o1 in Canvas and render HTML and React code there, reducing the friction between generating code and reviewing its output.
OpenAI strengthens Canvas’s direct comparison with Claude’s Artifacts, especially for users working across prose, code, and interactive prototypes.
Second-order effects
Anthropic faces greater pressure to differentiate Claude’s workspace on execution quality and workflow depth rather than the basic shared-workspace format.
The competitive unit shifts from the standalone model response to the surrounding surface: editing, previewing, and iterating within the assistant.
Third-order effects
If providers continue adding creation and preview tools to chat products, AI workspaces may consolidate writing and lightweight software-production tasks that previously required separate tools.
That shift could make interface integration and workflow continuity more important competitive moats than any single model capability, though user adoption will determine whether these surfaces become primary work environments.
The trend: Frontier-model vendors are competing to make the assistant a persistent work surface, not just a conversational endpoint.
Canvas update: today we're rolling out a few highly-requested updates to canvas in ChatGPT. ✅Canvas now works with OpenAI o1—Select o1 from the model picker and use the toolbox icon or the “/canvas” command ✅Canvas can render HTML & React code
More good @ChatGPTapp launches for you—shipping some top requests on a Friday! 🫡 1. Canvas can now render HTML/React, meaning you can make your own mini-apps using Canvas. 2. You can use o1 with Canvas (great for coding) 3. Canvas is available on ChatGPT desktop for Mac.
WOW! ChatGPT Canvas can now build React apps! Here is my first little Canvas app that lets me upload images and then it uses OpenAI vision API to explain them. Anyone can build apps in a matter of seconds. It can even call APIs! So cool! [video]
Canvas with o1 is available to Pro, Plus, and Team users. React/HTML code rendering is available to Pro, Plus, Team, and Free users. Both updates will roll out to Enterprise and Edu in a couple weeks.
One of the coolest examples I've seen: @edwinarbus instantly built a Windows 95 simulator in ChatGPT with canvas using @OpenAI o1—then started playing Minesweeper right there! It just worked! 💣
you can now use o1 to code in chatgpt canvas, and it renders html and react apps used it to quickly build a windows 95 sim! iterated on it within chatgpt—adding minesweeper, notepad files that can be saved and edited, even pointer traillssssss [video]
The new @OpenAI Canvas really needs a detailed manual What are the rules for loading external libraries? It seems to work for some of them Can it make outbound fetch() API calls? I've been getting all kinds of weird errors from it that I can't figure out [image]
The future is agentic... and generative. We are building towards highly personalized apps that are built just in time. Oh and /canvas is really nifty too.