Tools · 2026-10-31

Figma Weave: how to create images and videos with AI step by step and repeat the result

Figma Weave is a Figma tool that combines AI image and video generation and editing into one repeatable chain of steps that the whole team can use.

✨ What Figma Weave is

Figma Weave is a platform for creative teams where creating and editing images and videos with AI looks like a sequence of steps. Each step can be changed, and the whole chain can be reused.

Work in Weave is made of nodes. These are separate blocks on the canvas: for adding images and prompts, for generating content, and for editing. Connected nodes show how the result was made, and you can go back to any earlier step to change it.

📌 How it works in an example

Suppose you are making an advertising campaign for a hiking gear brand and start with a photo of a backpack. You connect this photo to two image-editing nodes: in one you ask to place the backpack in a forest, in the other on a mountain.

Then you can change the colors of the forest scene or turn the mountain scene into a short video. Both branches stay on the canvas, so you can see how each option would look in the campaign before choosing one.

🔹 How to repeat a workflow

A finished chain can be turned into a Weave tool. A colleague then does not set up the nodes; they just upload a photo and click Generate. The tool runs all the steps automatically and returns the final result.

For example, you can set up product photo processing once: remove the background, place the product on a branded backdrop, and crop it for a social media post. The same processing is then used for each new product launch. Publishing such tools to the community is already live. Examples include turning a moodboard into a realistic room mockup and previewing a new nail design.

🔸 Useful tips

Clear prompts give better results. Instead of “make the lighting better,” write “use soft, warm light, keeping the backpack’s color and shape unchanged.” If you need a specific style, add a reference image. If you have several references, write a separate prompt saying what to take from each, for example: “use the colors from the first image and the illustration style from the second.”

Make small fixes by hand instead of prompting again. This gives more control and does not spend credits on another generation. Also check the result at each step so an error does not carry through the chain. For example, check an edited image before using it to create a video.

Sources: 🔗 Figma Blog