The design platform has opened Code Layers to public beta, introduced Figma Specs to guide AI agents and expanded integrations with 20 additional tools, including India’s Sarvam AI.
By GAWAH News Desk | Technology
Figma is expanding its artificial intelligence capabilities to bring design, software code and AI agents closer together, in a move that could change how product teams build and test digital experiences.
The company announced a series of updates at Config India 2026 in Bengaluru on October 15, introducing new tools designed to help designers, developers and product managers collaborate within a shared workspace.
The announcements include the public beta rollout of Code Layers, a new Markdown-based documentation feature called Figma Specs, improvements to its AI integration ecosystem and deeper support for building data-driven websites.
The developments reflect a wider shift in the technology industry: AI tools are increasingly moving beyond generating text and images towards helping users create interactive prototypes and functional software.
Figma Code Layers: From static designs to working prototypes
One of the main announcements is the expansion of Code Layers into open beta.
Traditionally, designers create visual mock-ups that developers subsequently translate into working applications. Code Layers aims to reduce the distance between those stages by allowing teams to work with code-backed, interactive prototypes directly on Figma’s canvas.
Users can convert existing design frames into interactive prototypes, import code from GitHub, capture code from a live website or start building from a prompt.
Teams can then test different versions of an experience, inspect how it behaves and iterate before committing engineering resources to a finished product.
Figma says the feature is intended to help teams explore ideas and evaluate functionality earlier in the development process. The feature is rolling out across eligible plans, although availability may vary during the beta rollout. Users also need appropriate editing access to their design files.
Figma Specs gives AI agents a rulebook
Figma has also introduced Specs, a native Markdown-based environment where teams can document product requirements, design decisions and implementation details.
The idea is to give both people and AI agents a common reference for how a product should look and behave.
Teams can connect written guidelines with design elements, diagrams, code layers and design libraries. This provides additional context when an AI agent is asked to create or modify a component.
For example, a company could document its brand colours, typography, interface patterns and accessibility requirements. An AI agent could then refer to those guidelines while generating a new screen or revising an existing prototype.
Figma’s approach addresses a common problem with AI-assisted design: generated work may look convincing but fail to follow a company’s established design system or technical requirements.
The new documentation tools are intended to improve consistency, although AI-generated output will still require review and testing. Figma Specs is rolling out in open beta on eligible paid plans.
Payload integration brings websites and content together
Another part of the announcement concerns Payload, the open-source content management system acquired by Figma last year.
With Payload integrated into Figma Make, users can create websites and applications with a backend that includes content management, databases, media handling, APIs, authentication and an administrative interface.
This could help teams move from an initial prototype towards an application that works with real content and data, rather than stopping at a visual demonstration.
The development also places Figma in a more direct position within the website-building and application-development market, where design tools increasingly overlap with AI coding platforms and content management systems.
The practical implications will depend on how well these generated applications meet production requirements for performance, security, reliability and ongoing maintenance.
India’s Sarvam AI joins an expanded integration ecosystem
Figma also announced support for 20 additional partners in its Model Context Protocol ecosystem, including Indian AI company Sarvam AI.
Model Context Protocol, or MCP, is a framework that allows AI applications to connect with external tools and services through a standardised interface.
For design and development teams, these integrations can help AI agents work with information and tools outside a single application.
The expanded ecosystem includes tools such as OpenCode and other agent platforms, giving users more options for connecting their preferred AI workflows with Figma.
The addition of Sarvam is particularly relevant to India’s AI ecosystem, where domestic developers are building models and applications intended for local languages and use cases.
The announcement does not, by itself, mean that every integration supports every feature or is available to every user. Access depends on the relevant product, integration and account configuration.
Why Figma is moving beyond design
Figma’s latest announcements reflect a changing relationship between design and software development.
As AI makes it easier to generate code and interactive prototypes, teams can experiment with more ideas before deciding which ones are worth developing into finished products.
That could shorten some early development cycles and enable designers to participate more directly in building functional experiences.
It also creates new challenges. More generated code does not automatically translate into better software. Applications still need testing, accessibility checks, security reviews and engineering oversight before they can be reliably deployed.
The challenge for platforms such as Figma will be to make AI-assisted creation useful without sacrificing quality or control.
What this means for designers, developers and businesses
For designers, Code Layers could make it easier to test interactions without waiting for a separate engineering implementation.
Developers may benefit from earlier feedback on product behaviour and closer collaboration with design teams, although production code will still require technical review.
For startups and smaller businesses, the ability to build and evaluate interactive prototypes quickly could lower the initial cost of exploring a product idea. However, the cost and complexity of running a production application will depend on its infrastructure, traffic, integrations and security requirements.
The broader direction is clear: design platforms are increasingly becoming environments where people and AI agents can move between ideas, visual interfaces and code.
Figma’s latest features are an important step in that evolution, but their long-term value will depend on whether they help teams deliver better products, not simply generate more prototypes.