Figma's 3D transforms arrived earlier this year alongside Figma Motion, generative plugins and shaders built on the Figma agent, and new ways of working with existing materials. The 3D transforms effect is now in open beta, and the Figma agent is generally available. To show what the combination makes possible, Figma's Brand Studio team produced the feature's launch video entirely through these tools.

How the effects work

The 3D transform effect tilts, rotates and arranges text, images, vectors and frames in space. Adding the perspective effect to a parent frame pulls those elements into one shared spatial composition. Both effects are non-destructive: the underlying artwork stays editable, while text and fills render with the transformed layer. As with any Figma effect, 3D transforms can be combined with shaders, animated through Figma Motion, or explored further with the agent.

Typography as a spatial composition

Brand designer Deanna Gueli began with a single word, repeated and staggered in a black-and-white study. After working out an initial placement, she asked the Figma agent to repeat it at intervals along the Z axis, removing the need to hand-calculate each position. "There was a tango back and forth: The agent did math, and I did the creative stuff," she said.

To deepen the sense of depth, Gueli built geometry with the agent's help, placed the type animation onto a rectangular tube and turned the form into a component. Duplicating that component made it easy to test how many tubes should fill the frame, how far apart they should sit, and what happens when the whole arrangement rotates. Comparing compositions side by side on the canvas showed her which directions worked.

Keeping art direction open after animation

Components also kept the scene editable for the rest of the team. Brand designer Yeojin Shin could change colors and gradients through the components even after the scene was animated, and those adjustments carried into the animation. When the typography shot moved to a different point in the edit, the team could retune its palette to match the surrounding shots while watching the result in context.

Agent-built plugins for 3D layout

Brand motion designer Ben Hill wanted to give flat compositions depth by tilting, stacking and rotating frames, but arranging the cards by hand was slow. "So I had the Figma agent build a generative plugin to do the heavy lifting, arranging everything into a grid, a helix, a ring, or a randomized scatter," he said. With placement automated, Hill generated variations and compared compositions on the canvas, using placeholder shapes to settle depth, balance and movement before developing any artwork. He picked the randomized arrangement for its loose, expressive feel, then swapped the placeholders for photography, circles, lines and other graphic elements. Shaders added surface treatments, including light sweeps across the assets, so layout, art direction and animation could evolve together.

A vector flower, colored by shader

Hill next tested whether a set of vector shapes could become a dimensional illustration. He arranged the shapes in 3D space to form a flower, working in grayscale to resolve the form and composition before applying color through a gradient-map shader, which assigned colors to the artwork's light and dark values. He then placed the flower in a homepage concept for a fictional brand, alongside animated typography, logos and icons from brand motion designer Gilles Desmadrille. That context changed the motion: "The original flower animation had a more fluid, handheld camera style, but we wanted to make it a bit more simple and tied to a scrolling action," Hill said. The team pared the movement back and retimed it to work with the type animation.

A different order of decisions

Testing rough spatial compositions in motion before committing to final art direction reversed the team's usual creative process, since some ideas are hard to judge as static designs. The work also stayed collaborative throughout: brand designers adjusted colors and gradients through components while motion designers kept refining movement, with decisions made in context rather than by passing files between tools.

From canvas to code

3D transforms map directly to CSS. Selecting a 3D layer in Dev Mode exposes its exact CSS transform, including depth and rotation on every axis, ready to inspect and copy. Coding agents can also read and create 3D designs through the Figma MCP server.

3D transforms are in open beta for all Full seat users; the help center has documentation, and a playground file includes examples from Figma's product design and brand design teams.