Figma's 35-second video announcing the Code Layers open beta at Config India on October 15 had passed 6.7 million views on Figma's official X account as of October 26. A week later, Figma posted a separate 10-minute video showing four ways to use Code Layers. After trying it on his company's blog, Murat Bayral of the YouTube channel 02ui, who has more than 20 years of design experience, said, "this thing is basically changing Figma entirely".
You can bring a Figma frame, a GitHub repository, a folder on your computer, a webpage imported with the Chrome extension, or code sent by a coding agent connected to Figma through Figma MCP onto the canvas. You can interact with the screen, duplicate it to compare options side by side, and leave comments on it. Updated code can be pushed to a GitHub branch and submitted in a pull request for team review.
It is currently available to people with a Full seat and edit access to a Design file. According to Figma Help, creating and editing code layers on paid plans does not use AI credits through November 4. The feature is rolling out over several weeks, so some people on the same team may see the menu later than others. To import from GitHub, you must connect Figma to your GitHub account or organization. Only lightweight prototype code is supported, and importing can take more than 10 minutes. The Figma agent on the canvas creates and edits the code.
