Product listings
A marketplace photo on a plain white background is the case this tool exists for. The shadow is what separates a product photograph from a sticker.
Private image tools
Stop your cutout floating above the page
A real object occludes the light that would otherwise reach the surface beside it. That occlusion is most of what tells your eye the thing is sitting on something. Cut the background away and it goes with it, which is why a perfectly clean cutout on a white listing page still looks wrong.
This is not generation, and the distinction matters. The shadow is your own cutout outline, squashed toward its base, offset, blurred and faded. No model was trained, nothing is downloaded, and the same file with the same settings always produces the same bytes. Nothing on this page invents anything that was not already in your image.
Result proof
A transparent PNG that looks like it is standing on something
01
Input
A cutout: a PNG whose background has already been removed
02
Local pass
The alpha channel squashed toward the contact line, offset, blurred with three box passes and faded, then your original pixels composited on top
03
Result
A transparent PNG with a soft shadow under the subject
How to verify it: The result caption names the style, the strength, how much canvas was added and where the outline was read from. Open the result on a white page beside the original: the object should look the same and the space around its base should not. If the shadow looks detached from the object, the cutout has a soft fringe and the shadow is being grounded below where the object really ends, which is worth fixing in the cutout rather than here.
What people use this for
Slides and one-pagers
A cutout dropped into a deck reads as clip art without one. Contact grounds it; Drop gives it the lifted, graphic look a title slide often wants.
Composites onto a real scene
Run Cutout Match first so the colours agree, then add the shadow. They fix two independent halves of the same problem.
Stickers and cut-out artwork
The Drop preset keeps the silhouette intact and offsets it, which is the classic sticker outline rather than a physical shadow.
Limits worth knowing
- It needs a transparent cutout. An ordinary photograph has no outline in it, so the tool refuses rather than returning an unchanged file.
- A matte gives the silhouette and not the shape, so the shadow is a footprint of the outline. It cannot know which parts of the object are nearest the surface, and on something photographed from above it will not convince at any setting.
- It assumes one light, from roughly in front and above. Multiple light sources cast multiple shadows, and there is nothing in a cutout that would tell the tool how many there were.
- It does not shade the object itself. A real object sitting on a bright surface catches bounced light along its base, and adding that would mean changing your pixels rather than adding underneath them.
- The result is usually slightly larger than the file you supplied, because the offset and the blur need room. The exact amount added is printed with the result and the added area is transparent. The growth stops at 512 pixels a side, and on a large image at the Soft setting a shadow can be wider than that, in which case the result says it is clipped.
- There is no image-editor version. The editor keeps your photograph and your selection separately and combines them when it draws, so a shadow placed under it would be under opaque pixels and would never appear. This page is the tool.
- An image larger than the working limit this device can hold safely is scaled down to that limit first, so the result comes back at that size. The page says so on screen when it happens, and your own file is never changed.
How the local workflow runs
- Remove the background first, if you have not already. This tool needs the transparency.
- Choose the cutout. It stays on your device.
- Start with Contact. It is the preset that matches how most objects actually sit.
- Look at the base of the object rather than at the whole image: that is where a wrong shadow shows.
- Bring the strength down if it looks heavy, then download the PNG.
Frequently asked questions
Is this AI?
No, and it is worth being precise because a shadow that was not in your photograph now is. Every pixel of it is computed from the transparency your cutout already carries: the outline is squashed toward the point where the object meets the surface, moved slightly, blurred, and faded. Nothing was trained, nothing is fetched, and running it twice with the same settings produces identical files.
Why does my image come back slightly larger?
Because a shadow needs somewhere to go. It is offset below the subject and blurred outward, and if the subject already touches the bottom of the frame that would be clipped off. The page adds as much canvas as the blur and the offset need, on every side, up to a limit of 512 pixels, and the result caption tells you how much was added. The added area is transparent, so it costs nothing visually and it can be cropped away. On a very large image at the Soft setting the blur can reach further than that limit, and when it does the result says the shadow is clipped rather than pretending it is not.
What do the three styles actually change?
How far the silhouette is compressed toward the base, how far it moves, how soft it is and how dark. Contact keeps the shadow tight and dark under the object, which is what something resting directly on a surface casts. Soft spreads it out, as though the light were a window rather than a lamp. Drop does not compress it at all, so the shape reads as itself rather than as a footprint, and that one is a graphic effect rather than a physical one.
Why does it need a transparent PNG?
The shadow is built from the outline, and an ordinary photograph has no outline in it: every pixel is opaque, so there is nothing to say where the subject ends. Run the file through the background remover first and bring the transparent PNG back here. If you give it an opaque image the tool refuses and says so, rather than handing you back an unchanged file and reporting success.
Will it look right on any object?
No, and the reason is structural rather than a setting you can change. A matte tells the tool the silhouette and nothing about the shape, so it cannot know which parts of an object are nearest the surface. On a bottle, a box or a shoe photographed from roughly eye level that is close enough to convince. On something photographed from above, or something with a large overhang, the shadow will read as a footprint of the outline because that is exactly what it is.
Can I use it in the image editor?
No, and it is better to say so than to put a button there that does nothing. The editor does not hold a cutout: it keeps your original photograph and your selection as two separate things and combines them only when it draws. A shadow belongs underneath a cutout, and underneath an opaque photograph there is nothing to see, so the run would finish, report success and change not one pixel. It will get an editor surface when the editor can hand a tool a real cutout and take a larger canvas back. Until then this page is the tool.
Does my image leave my device?
No. The file is decoded by your browser and the shadow is built in the same tab. There is no upload, no server, and no model fetch.
Related tools
Background Remover
Where the transparency this tool needs comes from.
Cutout Match
The other half of why a composite reads as fake. Also no download.
Image Editor
Refine the cutout first: brush the mask, tune the edge, then bring the transparent PNG here.
Photo Check
Find out whether your photo will cut out cleanly before you start.
Ready to try it?
No upload, no model download, and your original is never changed.
Ground a cutout →