Introducing Visual Copilot 2.0: Make Figma designs real

Announcing Visual Copilot - Figma to production in half the time

Builder.io logo
Contact Sales
Platform
Developers
Contact Sales

Blog

Home

Resources

Blog

Forum

Github

Login

Signup

×

Visual CMS

Drag-and-drop visual editor and headless CMS for any tech stack

Theme Studio for Shopify

Build and optimize your Shopify-hosted storefront, no coding required

Resources

Blog

Get StartedLogin

‹ Back to blog

design to code

How to build an app dashboard with AI-powered Figma-to-code

February 12, 2025

Written By Josh Thomas

Using AI to generate code is table stakes these days, but what leaders inside organizations care about is finding tools and workflow that actually make their teams more efficient, output code they can trust, and improve overall employee happiness and productivity.

At Builder, we are focused on turning design into clean, working code, and combining that with component mapping so you can build and iterate on the AI’s output.

Today I want to walk you through how we're using AI to go from Figma designs straight to working code - specifically for building app dashboards and internal tools. We'll look at:

  • Importing designs from Figma into Builder
  • Generating responsive layouts
  • Mapping components to your codebase
  • Exporting production-ready React + Tailwind code

This post is a accompanying guide to a session from our Builder Labs series. Follow along with the original recording here!

Alright, Let's dive in!

The current workflow

Illustration of traditional Figma-to-code workflow without AI, showing a designer handing off figma file to a developer, with the developer's iterative process of studying design, creating layout, and implementing real data, plus a revision cycle involving both roles

Typically, turning Figma designs into working code involves a lot of manual work:

  1. Designers create mockups in Figma
  2. Developers study the designs
  3. Devs recreate layouts by hand in code
  4. Components get wired up with real data
  5. Back-and-forth to handle revisions

It's time-consuming and error-prone. Small design tweaks can mean big code changes.

There's got to be a better way, right?

Illustration of AI-powered Figma-to-code workflow, showing a Designer handing off the Figma file to Builder plugin, followed by an automated process where AI analyzes layouts, generates code, and maps to existing components, before syncing production-ready code to the Developer's editor”

Here's a high-level look at how our AI pipeline works:

  1. Start with a Figma design
  2. Use our Figma plugin to prep & import the design
  3. AI models analyze the layout and components
  4. Generate responsive React + Tailwind code
  5. Map to your existing component library
  6. Export production-ready code to your project

The whole process takes minutes instead of hours or days. Let's break it down step-by-step.

First, we install the Builder plugin in Figma. This lets us select our design and export it with a single click:

Preview of the Builder plugin in figma, with the export to code button highlighted. What is being exported is a file with two graphs on it, a line graph and a bar graph.

The plugin uses our proprietary AI models to:

  • Analyze the layout structure
  • Identify components and variants
  • Preserve naming and hierarchy
  • Prep everything for import

All of this happens locally in Figma - we don't send your design files to external servers.

Once imported, the design shows up in Builder's visual editor:

Here we can preview how it looks at different breakpoints. The AI has already made the layout responsive, but we can tweak things if needed.

This step is optional - you can go straight to code generation if you prefer. But it's handy for quick adjustments.

This is where things get really powerful. We can map Figma components to your actual React components:

For example, we can tell Builder that this Figma button:

Copy code1Button
2├─ Type: "Primary"
3└─ Size: "Large"

Should map to this React component:

Copy code1<Button variant="primary" size="lg">

Now anytime that Figma component appears, we'll use the proper React version.

You can set these mappings up once and reuse them across projects. We can even bundle them with your design system.

With everything set up, we hit "Generate Code" and Builder gets to work:

Not too shabby! This is production-quality code that's ready to drop into your app.

It uses a combination of our proprietary models and fine-tuned LLMs to create clean, idiomatic React code.

The output includes:

  • Responsive Tailwind classes
  • Proper component usage
  • TypeScript types
  • Accessibility attributes

The final step is getting this code into your project. Builder can push directly to your repo, or you can copy/paste if you prefer.

We're also testing a new CLI workflow that makes updates even smoother:

Copy code1builder develop --sync "Add dashboard to homepage"

This analyzes your existing code, figures out where to insert the new components, and handles the integration for you.

It's pretty slick - especially for iterative design changes.

So that's a quick tour of how we're using AI to streamline the Figma-to-code process. To recap the key benefits:

Mind map showing four key benefits of AI-powered design-to-code: Close to design, Consistent code, Easier design iterations, and Saves time
  • Massive time savings (minutes vs days)
  • Closer adherence to designs
  • Easier design iterations
  • More consistent code output

And don't forget, if you want to see a live demo that dives a bit deeper into demonstrating the details of the workflow above, you can check out our recent webinar

We're continually improving our models and adding new features. If you want to try it out, check out Builder.io.

Let me know what you think! Is this something you'd use in your workflow? Any features you'd like to see? I'm always eager to hear feedback.

Happy coding!

Precise Mode

Custom Instructions

Mapping Components

Code Gen CLI Beta

Join the private beta for the contextual CLI flow.

Join Here

Introducing Visual Copilot: convert Figma designs to high quality code in a single click.

Try Visual CopilotGet a demo

Share

Twitter
LinkedIn
Facebook
Hand written text that says "A drag and drop headless CMS?"

Introducing Visual Copilot:

A new AI model to turn Figma designs to high quality code using your components.

Try Visual CopilotGet a demo
Newsletter

Like our content?

Join Our Newsletter

Continue Reading
Design to Code8 MIN
Figma to Code with Windsurf and Visual Copilot
February 10, 2025
Web Development12 MIN
Replace your JavaScript Animation Library with View Transitions
February 3, 2025
design to code8 MIN
Figma to Code with Cursor and Visual Copilot
February 3, 2025