Figma: Easier Reference Images for AI Make Image & Edit Image
FigmaFigma improved the workflow for adding reference images to its AI-powered Make Image and Edit Image features on April 29, 2026. Previously, references could only be provided by uploading a file from disk; now designers can click an "Add reference" button on almost any canvas node, paste an image directly into the prompt box, or drag and drop files into it. The update applies consistently across all Figma products that support AI image generation: Design, Draw, Buzz, Slides, and FigJam.
Sources & Mentions
4 external resources covering this update
Streamlined Reference Inputs for AI Image Generation
Figma's AI image tools — Make Image and Edit Image — received a meaningful workflow improvement on April 29, 2026. The update directly addresses one of the most common friction points in AI-assisted image creation: providing reference visuals to guide the model's output.
The Previous Limitation
Until this update, adding a reference image to Make Image or Edit Image required uploading a file from the local file system. This workflow interrupted the canvas-centric experience Figma is built around — designers had to leave the tool, locate a suitable file on disk, and upload it, even when a perfectly appropriate reference was already present on the Figma canvas.
What Changed
Figma introduced three new methods for adding references:
Add reference from the canvas. A new "Add reference" button now appears on almost any node — images, vectors, frames, and components alike. Clicking it pins that canvas element as the reference for the active AI prompt. This keeps the entire workflow on-canvas and allows designers to reference existing design assets, illustrations, or photographs already in the file without any file system interaction.
Copy/paste into the prompt box. Designers can copy any image from the canvas or from an external source and paste it directly into the Make Image or Edit Image prompt input. This works for content sourced both inside and outside Figma.
Drag and drop into the prompt box. Files can be dragged from outside Figma and dropped directly into the prompt input, preserving the original upload workflow in a more discoverable and faster form.
Where It Works
All three methods are available across every Figma product that supports AI image generation: Figma Design, Figma Draw, Figma Buzz, Figma Slides, and FigJam. The consistent rollout means designers working across different Figma surfaces get the same improved experience regardless of which product they are in.
Why It Matters
AI image generation quality improves significantly when the model receives a visual reference alongside the text prompt — it anchors style, composition, and content to something specific rather than leaving interpretation entirely to the model. The previous upload-only friction likely discouraged many designers from using references at all. By making reference addition a one-click action on any canvas element, Figma removes that barrier and makes guided, higher-quality AI image output the path of least resistance.