ImgToLayers
Split a flat image into editable layers
About ImgToLayers
ImgToLayers does one job that sounds small until you have needed it: it takes a flat image and gives you back its parts. You upload a JPG, PNG, WebP, GIF or AVIF, choose how many layers you want, and it returns the subject, background, shadow, text and smaller details as separate raster layers that stay aligned to the original canvas. The default export is a layered PSD that opens in Photoshop with every layer where it sat in the source image, and the same layers can also be downloaded as transparent PNG or WebP files, one by one or as a grouped ZIP for Figma.
The problem it addresses is the one that makes flattened images frustrating. A product shot from a supplier, a poster someone exported, a screenshot of a design whose source file is gone: all of them are finished in a way that resists changing. Background removers solve one slice of that and throw the rest away, including the shadow that made the subject look real. Keeping the shadow as its own layer is usually the difference between a composite that looks deliberate and one that looks pasted on. The site is also unusually plain about what it cannot do. PSD export contains raster layers only, so it does not recover live text, vectors, smart objects, hidden pixels or original masks from a flat image, and the background setting is described as guidance to the model rather than a guarantee. Under the hood the workflow runs on the fal-ai/qwen-image-layered model. Every job is paid for in credits that scale with layer count: 4 layers costs 10 credits, 5 costs 13, 8 costs 20 and 10 costs 28. New accounts get 10 free starter credits, which is enough for exactly one 4-layer job, and you have to sign in before uploading anything.
Sign in and upload one image of up to 20MB. Pick the layer count, which is both the cost lever and the quality one: a 4-layer split is the cheapest, while 8 and 10 layers ask the model to separate finer detail and cost more. Choose the output format (layered PSD, PNG or WebP) and how the background should be handled: keep the canvas, split it into its own plate, or make it transparent. The app checks your sign-in and credit balance before the model runs, and asks you to buy credits if the balance is too low for the layer count you picked. When the job finishes you review the original and every separated layer before downloading. The PSD is assembled in your browser from the generated layers. Successful jobs stay in Settings, Layer history, so you can reopen them in the Layer Workspace and download the ZIP or PSD again later. If a generation fails or is cancelled, the credits are refunded when the job is next checked, and a failed row in the history shows the refund.
- •Layer separation, not just a cutout - subject, background, shadow, text, props and details come out as distinct layers instead of one masked subject
- •Layered PSD export - aligned raster layers keep their canvas position and stack order, so the file opens in Photoshop looking like the original image, already split
- •Transparent PNG, WebP and ZIP packs - download single layers or a grouped ZIP that imports cleanly into Figma, with no proprietary format
- •Background modes - keep the canvas, separate it as a plate, or make it transparent, depending on what you plan to do next
- •Pick your layer count - 4, 5, 8 or 10 layers per image, so simple jobs stay cheap and complex scenes get a deeper split
- •Credits refunded on failure - failed or cancelled jobs return their credits, and Layer history keeps every successful result
- •Focused tool pages - product photo layers, logo to layers, poster graphics, character separation and image to PSD each get their own page, all sharing one generator and one credit balance
Ecommerce and marketplace sellers will use it most. Product photos arrive flattened from manufacturers, and lifting the product with its shadow intact and dropping it onto a new background or ad variant is most of the work. Print-on-demand sellers can split artwork, mockup and shadow so one source image serves thumbnails, listings and ads. Designers handed a JPEG with no source file get a layered PSD to clean up instead of a rebuild from scratch, and UI designers can split screenshots into cards, charts and panels for a quick redesign reference. AI artists and animators can separate characters, props and background art for compositing or parallax tests. The best results come from clean images with visible subjects, clear edges and decent contrast. It is not the right tool if you only need the background gone, since a dedicated remover does that one job, often for free. It is also not a way to recover an original design file: text comes back as pixels, not editable type. The site says uploads are not used to train a model, but deletion and retention controls are still being built, so avoid uploading anything confidential for now.
Pricing
$12.90 - $59.90/mo
- Starter$12.90/mo
- Pro$29.90/mo
- Studio$59.90/mo
- Starter Pack$19.90 once
- Creator Pack$39.90 once
- Studio Pack$79.90 once
From the vendor pricing page, 2026-10-21
