Figma launches Specs and public beta code layers

Figma has announced a suite of new features designed to give developers and designers more control over AI-generated output, headlined by the launch of Figma Specs and the public beta of its code layers. According to TechCrunch, these updates aim to solve the non-deterministic nature of AI agents by providing a structured reference for design rules and code libraries. The announcement marks a significant shift in how Figma positions itself as a platform for software teams, moving beyond static design files to become an active environment for defining how AI agents should build and ship products.

The Details

The centerpiece of the update is Figma Specs, a new Markdown-based editor that allows users to define their design language in a simple, text-based format. This document serves as a centralized reference for both human teams and AI agents working within the Figma ecosystem. Unlike traditional design guidelines that might be buried in documentation, Specs can exist as a standalone file or co-exist directly on the canvas. Users can reference other Figma elements, such as canvas components, embed FigJam diagrams, and link to code layers, creating a rich context for the AI to interpret.

Simultaneously, Figma has moved its code layers feature from private beta to public beta. Code layers allow users to add actual coding to designs on the canvas, enabling more interactive prototypes. This feature, first announced in June, allows users to ask the Figma agent to turn a design into code or modify specific parts of the code to iterate on a prototype. By making this public, Figma is inviting a broader user base to test the integration of code and design within the same workspace.

Additionally, Figma is expanding its Model Context Protocol (MCP) support to over 20 new tools, including Open Code, Dia, Grok Bot, and Indian AI startup Sarvam. This expansion allows third-party agents to better adhere to brand design rules. The company also announced deeper integration with Payload, the open-source headless CMS it acquired last year. Payload is now integrated with Figma Make, a product for generating dynamic websites, allowing users to directly manage content. Figma noted that Payload’s downloads have increased sixfold in the last 12 months, indicating strong adoption of this headless CMS approach.

Context

Figma’s strategy reflects a broader industry trend where design tools are evolving into development environments. Last year, Figma acquired Payload to strengthen its headless CMS capabilities, and this integration with Figma Make represents the maturation of that acquisition. The company is also investing heavily in its presence in India, opening its first office there late last year. Figma stated that India is its biggest market outside the United States, with the platform used by 40% of India’s top 100 companies listed on the Bombay Stock Exchange. The launch of these features during Figma’s first Config event in India underscores the company’s focus on this growing market.

The push for deterministic AI outcomes is a response to the limitations of current generative AI. While AI can generate code and designs rapidly, it often lacks the specific context of a company’s brand and technical standards. By introducing Specs and code layers, Figma is attempting to create a feedback loop where the AI is constrained by explicit rules, leading to more predictable and usable outputs.

What It Means for Developers

For developers, the introduction of Figma Specs means that design guidelines are no longer just for designers. The Markdown format makes it accessible for engineers to read and understand the design logic, potentially reducing the back-and-forth between design and engineering teams. The public beta of code layers allows developers to prototype with real code within the design tool, which can speed up the early stages of product development. However, developers should be cautious about the non-deterministic nature of AI agents. Even with Specs, the AI may still produce unexpected results, so manual review and testing remain essential. The integration with Payload and Figma Make offers a new pathway for building dynamic websites directly from design files, which could streamline the workflow for full-stack developers.

What to Watch

• Adoption of Figma Specs: Will teams actually use the Markdown editor to define their design systems, or will it remain a niche feature?
• Stability of Code Layers: As the feature moves to public beta, how reliable is the code generation, and how well does it integrate with existing development workflows?
• MCP Ecosystem Growth: The expansion to 20+ MCP tools suggests a growing ecosystem of third-party integrations. Which of these tools will become standard in Figma workflows?
• India Market Impact: With India being Figma’s largest market outside the US, how will these new features affect adoption and usage in the region?