OpenAI Codex and Figma Launch Two-Way Code-to-Design Workflow

OpenAI and Figma have launched a direct integration that connects Codex to Figma's design platform, letting product builders generate Figma designs from code and implement Figma files back into code. The release, announced by the two companies, turns what has traditionally been a one-way handoff between design and engineering into a roundtrip workflow.

The integration runs through the Figma MCP Server, built on MCP, an open-source standard that lets AI agents interface with external data sources, applications, and tools. Teams can pull details from Figma Design, Figma Make, or FigJam into Codex to implement them in code. In the other direction, users can turn UI from code into editable Figma designs, explore iterations on the canvas, and then bring the changes back into the codebase. Whether a product idea starts from a prompt, in code, or a design, the Figma MCP Server connects the pieces without losing context along the way, according to the announcement.

The stakes here are structural. As AI-assisted development compresses the distance between idea and working software, the boundary between design and engineering roles has become a contested frontier — and vendors that own both sides of that workflow stand to lock in more of the product-building pipeline.

Executives frame it as a role-softening shift

Figma's Chief Design Officer, Loredana Crisan, cast the integration against the backdrop of collapsing barriers to building software. "As the barriers for building software go down, the amount of software created will increase exponentially. It's no longer about whether you can build, but what you build and how it stands out," Crisan said. "With this integration, teams can build on their best ideas—not just their first idea—by combining the best of code with the creativity, collaboration, and craft that comes with Figma's infinite canvas."

Alexander Embiricos, Codex product lead at OpenAI, framed the change in terms of who the tool now serves. "This integration makes Codex powerful for a much broader range of builders and businesses because it doesn't assume you're 'a designer' or 'an engineer' first," Embiricos said. "Engineers can iterate visually without leaving their flow, and designers can work closer to real implementation without becoming full-time coders. The boundary between roles starts to soften because the system helps translate between intent and reality continuously."

Built on an existing partnership

The Codex-to-Figma integration is not the first tie between the two companies. Figma uses OpenAI models across its platform and was one of the first partners to launch a ChatGPT app in 2025. Figma has also deployed ChatGPT Enterprise across its organization to build AI fluency in its workforce.

For OpenAI, the integration adds momentum to Codex, its agentic coding product. Codex first launched as a CLI in April 2025 and now spans multiple surfaces: the terminal via CLI, an IDE extension, the web, and most recently a desktop app for macOS that supports multitasking with agents and automating background workflows.

The adoption numbers the company cites are substantial. Over 1 million people use Codex weekly, and Codex usage has increased more than 400% since the start of the year. Users range from product designers and data scientists to engineers, researchers, and product builders.

Enterprise uptake includes Cisco, NVIDIA, Ramp, and Datadog, along with startups such as Harvey and Sierra.

Why it matters

The integration sits at the intersection of two converging trends: agentic coding tools that automate implementation, and design platforms racing to become AI-native. By connecting Codex directly to Figma Make and FigJam through MCP, OpenAI and Figma are positioning themselves as a single environment for the full build cycle — prototype to production application — rather than separate stops on a fragmented toolchain.

To use the integration, builders install the Figma MCP server directly in the Codex desktop application. Whether the roundtrip workflow meaningfully softens role boundaries, as Embiricos predicts, will depend on how quickly teams adopt it — but with Codex usage up more than 400% this year, the audience for it is already large.