Figma Revolutionizes Creativity: Unites Design, Code, and AI on One Canvas!

During its first significant product conference since going public, Figma took the opportunity to redefine the core functionalities of its tool. At Config 2026, its annual conference in San Francisco, Figma unveiled several new features that integrate design, code, creative materials, and agents within a single, intelligent fullstack creation canvas. This move extends the canvas’s previous opening to AI agents initiated in the spring.

Creative Materials Bridging Design to Code

The first category encompasses what Figma refers to as new creative materials, intended to expand the possibilities directly on the canvas. A key innovation is the introduction of code on the canvas: teams can now clone a repository, keep it active within Figma Design, and explore new directions with the Figma agent.

This functionality works both ways, as it is now possible to extract a code stream into editable design layers and then resynchronize changes back to the code. Essentially, design and development are based on the same foundation, eliminating the usual break between mockup and integration.

The three other materials continue this logic of creation directly on the canvas:

• Figma Motion: a native module for animations, transitions, and 3D scenes, which can be used through prompts, predefined styles, or on a timeline. Motion is connected to design systems and generates code ready for deployment.
• Shaders: prompt-generated custom visual effects (screening, pixelation, advanced blurs) powered by WebGPU, directly on the canvas without external tools.
• Workflows Figma Weave: over 20 Weave tools integrated into the canvas to convert complex AI workflows into simple functions. Figma describes this step as the first towards a full integration with Weave (formerly Weavy), expected by the end of the year.

Intelligent Tools for Teams and Agents

The second category relates to agents, which Figma aims to make more autonomous and better contextualized:

• Agent skills and expanded context: repetitive tasks can be transformed into skills shareable with the entire team, and the agent can be enhanced through third-party connectors, web searches, and attachments.
• Creation of plugins with the agent: it is now possible to build custom, reusable plugins from prompts, without installation or technical skills.
• Agents in FigJam and Slides: the agent extends to other products, to create and analyze diagrams in FigJam, and generate content, make bulk modifications, or synthesize feedback in Figma Slides.

A Strategic Approach to Merging Design and Code

These announcements extend a strategy that began earlier in the year. After integrating Claude Code and Codex in February, Figma opened its canvas to third-party agents in March through the use_figma tool, before deploying its own native AI agent in May. “Too often, the tools we have for creation force us to make false choices between design and code, between humans and agents,” stated Dylan Field, co-founder and CEO of Figma, in the press release.

This acceleration comes in a competitive environment. Google is advancing its design tool Stitch, now equipped with a free multiplayer agent, while Canva and Adobe are also enhancing their creative assistants. It remains to be seen whether this expanded canvas will be enough to keep Figma ahead in a market where the line between design and development is increasingly blurring.