Skip to main content

Edit multiple photos at once on canvas

Apply consistent edits across dozens of images with Figma's agent—adjusting exposure, removing backgrounds, or retouching details in bulk. No exporting to a separate editor, no rebuilding your layout afterward.

  • AirBnb
  • Atlassian
  • Dropbox
  • Duolingo
  • GitHub
  • Microsoft
  • Netflix
  • The New York Times
  • Pentagram
  • Slack
  • Stripe
  • Zoom

Edit at scale without losing precision

Figma's agent handles repetitive photo edits across your entire file, so you keep your focus on the design decisions that matter.

Edit where your designs already live

Figma's agent works as a batch photo editor right on your canvas, applying consistent changes to multiple images without exporting to another tool and importing each result back one by one.

Cut the busywork, stay in flow

Hand off repetitive edits—background removal, color correction, exposure adjustments—across your entire image set and keep your focus on the creative work that really matters.

Keep your team in sync

Batch edits happen in the same file your team works in, so everyone sees the latest versions and can build on them immediately. Nothing gets lost in exports or version confusion.

The features behind the agent

A closer look at what sets Figma's agent apart, built for the way design teams actually work.

  • Screenshot showing Figma AI agent connected to a component library, auto-applying brand tokens to a designScreenshot showing Figma AI agent connected to a component library, auto-applying brand tokens to a design

    Connect to a design library

    Generate and refine using your actual components, tokens, and variables—not placeholder UI you'll have to swap out later.

  • Screenshot of multiple simultaneous AI design prompts running in FigmaScreenshot of multiple simultaneous AI design prompts running in Figma

    Parallel prompting

    Run multiple prompts at once so you're evaluating directions in parallel, not burning time on each one in sequence.

  • Screenshot of a shared Figma AI conversation thread with design commentsScreenshot of a shared Figma AI conversation thread with design comments

    Shareable threads

    Share a thread and your team sees the full prompt, context, and iteration history behind any design. No separate handoff needed.

  • Screenshot showing Figma AI agent skill options for automating design workflowsScreenshot showing Figma AI agent skill options for automating design workflows

    Skills

    Package your team's workflows and conventions into reusable /commands anyone can run on demand.

Figma feels like a natural extension of my hand. I do my design ideation on the canvas, and now the agent takes it the last mile. With my design libraries connected, I can move from design to code faster than ever before.

Andy Madrick
Product Designer, Notion

Notion logo

Get started with Figma's agent

  • Step 1: Select your images

    Open your file in Figma Design and select multiple images on your canvas. Group them in a frame or select across frames—the agent handles both.

  • Step 2: Describe the batch edit

    Tell the agent what to change across all selected images—remove backgrounds, adjust brightness, apply a color grade. Add a reference image if it helps.

  • Step 3: Review the results

    Check the edits across your images on the canvas. A checkmark appears when processing completes. Refine in the same thread or try a parallel direction.

  • Step 4: Keep refining

    Make changes directly in your file based on the feedback. Run another review to validate improvements or explore new directions.

Become an agent expert

  • The Figma design agent is here

    Watch Figma's AI Agent in action and see how to put it to work in your own workflows.

    Read more
  • Figma agent playground intro screen with a 'Start' button and a sample chat prompt: 'Hey Figma, can you create different layouts for my reading app?Figma agent playground intro screen with a 'Start' button and a sample chat prompt: 'Hey Figma, can you create different layouts for my reading app?

    Work with the Figma agent

    Hands-on playground to explore Figma agent's capabilities — from bulk edits to feedback implementation — with step-by-step guidance built right in.

    Get started
  • How to use Figma agent in your design workflows

    Learn how to use Figma agent in your design workflows with a step-by-step tutorial from the Figma team.

    Watch now

FAQ

Figma's agent generates vector graphics directly on your design canvas. You describe what you need—icons, illustrations, UI elements—and the agent creates scalable, editable vectors using your design system's styles and colors. Everything stays in your file, ready to refine or use immediately.

Yes. Figma's agent creates vector images from text prompts right in Figma Design. The output is fully editable—you can adjust paths, colors, and shapes just like any other vector in your file. There's no conversion step or separate tool required.

Yes. Figma's agent can resize batches of images to consistent dimensions based on your prompt. Select your images, describe the target size or aspect ratio, and the agent applies the change across all of them. The resized images remain editable on your canvas, so you can make further adjustments if needed.

Figma's agent references your connected design library when generating vectors. This means outputs use your actual colors, styles, and tokens from the start. You spend less time restyling generic assets to match your brand and more time designing.

Figma's agent is available on all paid Figma plans. You'll find it in the left side rail in Figma Design, or open it with CMD + Enter. During the current open beta, it doesn't consume AI credits—so it's a good time to explore, experiment, and share your feedback with us.