Figma just gave designers AI shaders and a timeline while its stock sits 79% below its IPO
Figma launched native motion design, AI-generated shaders, and code layers at Config 2026. Here's what changes for designers and why the timing matters.
Figma unveiled Figma Motion, AI shader tools, and code layers at the opening keynote of Config 2026, held June 24 at San Francisco's Moscone Center.
- A real animation timeline, finally – Figma Motion brings keyframed animation directly into Figma Design, including every shader property.
- Shaders without graphics code – Designers can describe an effect or upload a reference image, and Figma's agent generates a working WebGPU shader.
- Launched as the stock dropped 79% – Figma is leaning into AI-native canvas tools while its share price sits far below its July 2025 IPO level.
Figma's stock has fallen roughly 79% since its July 2025 IPO, trading around $24, and on June 24, the company responded by handing designers a feature set that would have been unthinkable a year ago: native motion design, AI-generated shaders, and code that lives directly on the canvas. The announcements came at the opening keynote of Figma's Config 2026 conference at San Francisco's Moscone Center, where in-person tickets sold out ahead of the three-day event.
The headline feature is Figma Motion, which puts a real animation timeline directly inside Figma Design for the first time. Every property a shader exposes can now be keyframed on the motion timeline, something Figma's traditionally limited set of animatable properties never allowed. Product designer Adanna Onuekwusi, who tested the feature ahead of launch, said having everything in one space lets her systematize motion work and publish it as a shared library instead of rebuilding animations from scratch on every project.
The second major addition solves a problem most designers have quietly lived with for years: shaders. The new shader feature uses WebGPU for effects like dithering, pixelation, frosted glass, and polished chrome, visual effects that used to require actual graphics programming knowledge. Now a designer describes what they want, or hands the Figma agent a reference image, and the system generates a working shader with adjustable parameters already wired up.
"If a teammate can easily inspect and translate motion with the formats that we need, it reduces the gap between intent and shipping." - Alexandra Pereira, Senior Product Designer, Atlassian
What this means for the day-to-day designer: animation and shader work that used to require routing through After Effects or a separate shader tool, then converting the output into something a developer could implement, now happens inside the same file as the components and variables a team already maintains. The result exports to CSS and connects through Figma's MCP server, so the spatial and motion design a team builds is the same thing that ends up in code, not a reinterpretation of it.
Figma also launched code layers, letting designers convert any design layer into interactive code with a single click or prompt, treating code "like any other design material". Teams can clone repositories and extract flows from code directly into design layers for testing, collapsing a handoff step that has defined the design-to-development relationship for over a decade. Chief product officer Yuhki Yamashita framed it as changing collaboration "not just with designers, but also with engineers and PMs."
The timing isn't subtle. Anthropic's Claude Design tool can generate a clickable interface from a single prompt in minutes, a direct challenge to the premise that a digital product has to start with a Figma file at all. The irony is structural: Figma's own AI features run on rented models from OpenAI, Anthropic, and Google, the same companies building tools that threaten to route around Figma entirely. That dependency has a real cost, Figma's gross margin fell from about 92% to 86% over 2025, a decline the company ties directly to AI inference costs.
Figma's bet is that bringing AI generation onto its own canvas, rather than letting it happen around the canvas in a separate tool, keeps designers from ever needing to leave. Whether that's enough depends on whether teams treat Figma as the place design happens, or just one stop in a workflow that increasingly starts with a text prompt somewhere else. Interactive shaders are already promised as a next step, Figma says it's still tuning performance before shipping them.
