AI product photography in Figma works in two steps. Figma edits the images you already have: Remove background, Expand, Boost resolution, Erase and a background swap all run on canvas without a plugin. Shots that depend on your physical product are generated from your photo first, then placed on the canvas. That means new camera angles of the same object, a garment on a model, or product video.
You have one product photo, a PDP layout open in Figma, and six placements to fill by Friday. The photo is a three-quarter front view. The layout needs a top-down, a detail crop, and the same product on a model.
No amount of editing gets you there, because the shots you need were never taken. And next month it is 40 more products and the same six placements, which is the part hand-built canvas work does not survive.
So the work splits into two jobs. Figma finishes the pictures you have. Something else has to make the pictures you do not have. This covers what Figma’s AI does with images today, where its published tool list stops, the DesignerBox Figma plugin, the MCP route for repeating the same job, and what each shot costs before you run it.
Key Takeaways
- Figma removes backgrounds natively. Remove background, Expand images, Boost resolution, Isolate and erase objects, and Make and edit images all ship in Figma Design (help.figma.com, October 2026). A plugin that only removes backgrounds now overlaps with a tool built into Figma.
- Weave tools add Replace background and Change lighting on Professional plans and above. They were in open beta and used no AI credits until 6 October 2026. Since that date they are generally available, and each run uses Figma AI credits (help.figma.com, October 2026).
- Figma’s help pages for its AI tools and its Weave tools name no tool for virtual try-on or for a new camera angle of the same physical product (help.figma.com, October 2026). Both need a model that starts from your product photo.
- The DesignerBox Figma plugin puts generation on the canvas. Add it from the Figma Community listing.
- MCP covers the repeatable route. The DesignerBox MCP server is live with 75 tools.
- Images reach a Figma file through upload_assets, not through the write-to-canvas tool, which documents no image support yet (developers.figma.com, October 2026).
- The image model in the step sets the rate, and the cost is shown before the run. There is a free plan.
What Figma’s AI does for product photos today
Figma got good at images. Its AI image editor page now describes the work through an on-canvas agent: “Remove backgrounds, swap objects, retouch details, and refine visuals with AI, directly in Figma Design” (figma.com/solutions/ai-image-editor, October 2026). Figma’s help center names the image tools behind it: Make and edit images, Remove background, Boost resolution, Expand images, and Isolate and erase objects (help.figma.com, October 2026). Figma changes this surface often, so check the page in the month you plan around it. Every AI tool built into Figma lists the current set, with the plans and seats that get each one.
| Tool | What it does |
|---|---|
| Isolate and erase objects | Select an area to make it a new image layer, or erase it from the image |
| Expand images | Extend an image past its original borders |
| Boost resolution | Raise the sharpness and clarity of a low-resolution image |
| Remove background | Isolate the main subject |
| Make and edit images | Make or edit an image from a written prompt |
Weave tools in Figma Design sit on top of that, adding Replace background, Change lighting, Text to vector illustration, and Apply color palette, with Generate mockup and Transfer style also listed. They have been in Figma Design since 24 June 2026, on Professional, Organization and Enterprise plans, for people with edit access to the file. Weave tools in Figma Design were in open beta and used no AI credits until 6 October 2026. They are now generally available, and each run uses Figma AI credits (help.figma.com, October 2026).
Read that list twice if you were about to pay for a background remover. Figma does it on canvas.
Where Figma’s published tools stop
Most of the tools above start from an image you already have. They edit that image, extend it or restyle it.
Expand images invents more background. Boost resolution sharpens what is there. Make and edit images changes the image you uploaded, or makes a new one from a prompt. Weave tools go further: Figma’s launch post lists product shoots among more than 20 AI image tasks (Figma blog, June 2026). None of the named tools is built to show you the back of your shoe, because the back of the shoe is not in the file.
Two help pages list the tools by name: the AI tools in Figma Design and the Weave tools in Figma. As of October 2026, neither page names a tool for these jobs (help.figma.com, October 2026):
- A new camera angle of the same physical product. Not a similar product. Yours, from a side the camera never saw.
- A garment on a human model. Virtual try-on is not on either list.
- Product video. Figma Weave, the separate product, lists image-to-video animation (figma.com, October 2026). The tools on these two lists are image tools.
That is scope, not criticism. Figma is a design tool that learned to edit images well. Generating your product from angles that were never shot is a different job, and it needs a model that starts from your actual product photo.
The DesignerBox Figma plugin
The plugin brings DesignerBox generation to the canvas you are already working in. Add it from the Figma Community, then open it from the Plugins menu and sign in with your DesignerBox account. The plugin page covers what it does.
It covers work that Figma’s own image tools do not: new camera angles of your actual product and product stills generated from one source photo, landing in the file rather than in a downloads folder. A garment on a model is a DesignerBox template, and it is not one of the plugin’s six tools.
Six tools ship in the Apps tab: generate image, generate video, remove background, expand, product ad, and photo angles. Each one is broken down in the Figma AI plugin guide.
The plugin is the right tool for one file open in front of you. When the same six shots are due for 40 products, the work belongs in a saved workflow instead, and MCP is how you drive one from outside the browser.
The repeatable route: MCP
MCP, the Model Context Protocol, is an open standard for connecting AI applications to external systems. It was created at Anthropic and open-sourced in November 2024 (anthropic.com, November 2024), then donated to the Linux Foundation’s Agentic AI Foundation in December 2025 (anthropic.com, December 2025). Claude, ChatGPT and Cursor all support remote MCP servers over HTTPS (claude.com, developers.openai.com, cursor.com, October 2026).
The DesignerBox MCP server is live, exposing 75 tools, including image and video generation, avatars, brand profiles, and saved workflows. Your AI client drives the same models the web app does, and it can rerun a workflow you already approved rather than rebuilding the prompt each time. Workflow builders covers how that gets set up, and how to make your AI agent creative covers the connection from the client side.
That gets you the shot. Landing it on the canvas is the second half.
Figma shipped its own on-canvas design agent in May 2026 (help.figma.com, May 2026), and the agent left beta on 6 October 2026 (help.figma.com, October 2026). For how in-canvas assistants compare with the autonomous kind, see what to hand an AI creative agent and what to keep.
Getting the image onto the Figma canvas
Figma runs its own remote MCP server (developers.figma.com, October 2026), and two of its tools are easy to confuse.
Figma exposes two different tools, and they do not have the same powers:
- use_figma, the general write-to-canvas tool, lists “No assets (image) support yet” among its current limits. If you have seen a gray box where your image should be, that is why (developers.figma.com, October 2026).
- upload_assets does handle images. Figma documents it as uploading PNG, JPG, GIF and WebP assets into a file, up to 10MB each. Give it a node URL and the image lands as a fill on that node. Give it none and Figma creates new frames with the images as fills (developers.figma.com, October 2026).
Images go through upload_assets, not through the tool that writes your designs. Two caveats before you build a pipeline on it:
- Figma lets only clients in its MCP catalog connect, and names VS Code, Cursor and Claude Code as examples. Its write to canvas page also lists Claude Desktop and Codex among the supported clients (developers.figma.com, October 2026).
- Write to canvas is free during the beta, and Figma states it will become a usage-based paid feature (help.figma.com, October 2026).
If that is more plumbing than the job deserves, the short path is unglamorous and reliable. Generate in DesignerBox, drag the file into Figma. It takes seconds, and nothing in it is in beta.
The workflow, start to finish
- Shoot one accurate photo of the real product. Everything downstream derives from it, so this frame is the one that has to be right. Even lighting, true color, no clutter.
- Generate the shots the layout needs. The Photo angles app turns that single frame into several views. The Clothes on a model app puts a garment on a model. It starts on the Premium plan.
- Pick the model per shot if it matters. The model catalog runs image and video models from several providers, and each one carries its own credit price. Which model suits which product shot covers the choice.
- Bring it into Figma. Use the plugin, drag the file in, or drive upload_assets from a catalog client.
- Finish on canvas. Expand to the aspect ratio each placement wants, boost resolution, erase what the layout does not need. Figma does this part well.
- Save the run once it is right. The settings that produced an approved shot become a workflow you run again for the next product, so row two hundred gets the same treatment as row one.
Step 5 shows the split. DesignerBox produces shots that do not exist, and Figma finishes the ones you have. Step 6 is what turns a good afternoon into a process, which marketing teams call creative automation.
What each shot costs
There is no single credit price for an image. The image model you pick for the step sets the rate, and the spread between the cheapest and the most expensive model is wide enough to change a budget. A cheaper model and the default model are not close on the same six shots.
What matters more than the rate is that you see it first. The number is on screen before you press Run, so a six-shot PDP set can be priced before any of it runs rather than read off the bill afterwards.
Video costs more than stills, and the model on the step decides how much. An 8-second clip costs 40 to 560 credits, depending on the model. AI video starts on the Premium plan. Pick the model that fits the job.
Three gates worth knowing before you plan around them:
- AI video and virtual try-on start on the Premium plan. They are not on Basic or Pro.
- Uploading your own photos and the commercial license start on the Pro plan. The free plan runs on sample products. If the image runs as a paid ad, Pro is the floor.
- Every plan below Ultra is one seat. Team features, shared brand kits and white label are on the Ultra plan, so a design team sharing one brand setup needs it.
Plans and credits are on the pricing page. Check it before a campaign, not after.
Against what a product photoshoot costs, the arithmetic is not close. The trade is that you still need one accurate photo of the real thing.
Anyone can make an AI picture. Making hundreds that still look like your brand is the hard part. DesignerBox keeps the pictures, the brand rules, the editors, Assets and batch in one account, so a Figma file and a catalog run use the same brand. See how the four stages fit together.
The template library is where to start when you would rather open a finished job than a blank prompt.
FAQ
Can Figma remove an image background without a plugin?
Yes. Remove background ships natively in Figma Design, alongside Expand images, Boost resolution, Isolate and erase objects, and Make and edit images. These standalone AI tools need a paid plan and a Full seat (help.figma.com, October 2026). Weave tools add Replace background on Professional plans and above. They were in open beta and used no AI credits until 6 October 2026. Since that date they are generally available, and each run uses Figma AI credits (help.figma.com, October 2026).
Does Figma have virtual try-on?
Virtual try-on does not appear in Figma’s published tool lists as of October 2026 (help.figma.com, October 2026). Figma’s image tools edit the image you already have. Putting a garment on a model means generating something the photo does not contain, which is a different capability. DesignerBox includes try-on from the Premium tier.
What does an AI product image cost in credits?
It depends on the model. The image model you pick for the step sets the rate, and the cost is shown before the run, so you can price a shot list instead of reading the bill after it. There is a free plan, and it runs on sample products. Uploading your own photos starts on the Pro plan.
How do I get the DesignerBox Figma plugin?
Add it from the DesignerBox listing in the Figma Community, then open it from the Plugins menu and sign in. It puts DesignerBox generation on the Figma canvas: new angles of your product, product ads and stills built from one source photo. For repeatable work, the DesignerBox MCP server exposes 75 tools to clients that support remote MCP over HTTPS.
How do I get an AI-generated image into a Figma file?
Two ways. Drag the exported file onto the canvas, which needs no setup. Or use Figma’s MCP server: upload_assets accepts PNG, JPG, GIF and WebP up to 10MB and places them as a fill on a node or as new frames (developers.figma.com, October 2026). The general write-to-canvas tool does not handle images yet, so the upload path is the one to use.
Which AI clients can connect to Figma’s MCP server?
Only clients in Figma’s MCP catalog can connect. Figma names VS Code, Cursor and Claude Code as examples, and its write to canvas page also lists Claude Desktop and Codex (developers.figma.com, October 2026). The DesignerBox MCP server has no such catalog. It is built for clients that support remote MCP over HTTPS, including Claude, ChatGPT and Cursor.
Can I use AI product images in paid ads?
The commercial license starts on the Pro plan. Verify the current terms on the live pricing page before a campaign, and check each ad platform’s own AI disclosure rules separately, since those are set by the platform and change independently.
Sources
- Figma’s AI image editor page: (figma.com/solutions/ai-image-editor, October 2026)
- Figma’s AI tools in Figma Design, the tool names and the paid plan and Full seat requirement: (help.figma.com/hc/en-us/articles/23870272542231, October 2026)
- Weave tools in Figma Design, the named tools, plan eligibility, the open beta and general availability from 6 October 2026: (help.figma.com, read 7 October 2026)
- Product shoots among more than 20 AI image tasks, Figma blog, Connecting Figma and Weave, 24 June 2026: (figma.com/blog/connecting-figma-and-weave, October 2026)
- Figma Weave, the separate product, and image-to-video animation: (figma.com/solutions/figma-ai-tool-weave, October 2026)
- Figma’s MCP server, upload_assets formats and 10MB limit, the “No assets (image) support yet” limit on write to canvas, the supported clients and the catalog rule: (developers.figma.com, October 2026)
- Figma’s MCP server guide, free during the beta and later a usage-based paid feature: (help.figma.com/hc/en-us/articles/32132100833559, October 2026)
- Figma’s on-canvas design agent, shipped May 2026 and out of beta on 6 October 2026: (help.figma.com, May 2026 and read 7 October 2026)
- MCP created at Anthropic and open-sourced: (anthropic.com, November 2024)
- MCP donated to the Linux Foundation’s Agentic AI Foundation: (anthropic.com, December 2025)
- Remote MCP support over HTTPS in Claude, ChatGPT and Cursor: (claude.com, developers.openai.com, cursor.com, October 2026)
- DesignerBox Figma plugin listing: figma.com/community
- DesignerBox plans, credits and feature gating: designerbox.ai/pricing, October 2026
- DesignerBox model catalog and MCP server: designerbox.ai/models and designerbox.ai/mcp, October 2026
Figma capabilities re-checked on 2 October 2026 against help.figma.com, developers.figma.com and figma.com. The Weave tools’ general availability and plan rules were re-read on 7 October 2026. MCP facts verified against anthropic.com, claude.com, cursor.com and developers.openai.com. DesignerBox plan gates read from designerbox.ai in October 2026. This category changes monthly, so re-check any capability claim before relying on it. Individual results vary.
