AI cleanup

Erase unwanted objects with a brush

Paint over what should disappear. Magic eraser fills the selected area while preserving the surrounding scene.

Processed on our secure servers
Drop an image to erase an object
Paint over anything you want removed — Pro AI tool
or press Ctrl+V
PNG · JPG · WEBP · max 25.0 MB

How to remove an object from a photo

  1. Open the photo

    Drop a JPG, PNG or WebP up to 25 MB onto the page. It opens in the brush view with the original image at full resolution, so you can see exactly what you are masking.

  2. Paint over what should disappear

    Brush across the object, covering its shadow and any reflection as well as the object itself, and extend the mask a few pixels past its edges. A mask that stops at the outline leaves a faint ghost of the original behind.

  3. Run the eraser

    Press erase to send the image and your mask to a GPU worker. The model rebuilds the masked region from the surrounding scene rather than blurring or patching it. Each run costs 3 credits and takes a few seconds.

  4. Check the edges, then download

    Look at the boundary of the erased region at full size before you accept it. If a ghost or a mismatched texture is visible, widen the mask and run again. The result downloads as a lossless PNG named after your original with an -erased suffix.

The complete guide to removing objects from photos

What object removal actually does

Erasing something from a photograph is not deletion. An image is a fixed grid of pixels, and removing an object leaves a hole that has to be filled with something. The question is only what. The old answer was cloning: copy pixels from elsewhere in the image and blend them over the gap by hand, adjusting the source point constantly so the repetition does not become obvious. It works, and it is slow.

The modern answer is inpainting. You mark the region to discard, and a model generates entirely new pixels for it, conditioned on everything outside the mask. It has learned from a very large number of photographs what walls, pavements, foliage, sky and water tend to look like when they continue, so it can extend the background across the gap rather than copying a sample of it.

The practical difference is what happens at the boundary. A clone-stamp repair has to be blended at its edges or the seam shows. An inpainted region is generated with the surrounding pixels as context, so the transition is continuous by construction. The trade-off is control: the result is a prediction, and if the prediction is wrong your options are to re-mask and try again rather than to nudge it.

Masking well, which is most of the job

The quality of the result is decided almost entirely by the mask, and the most common mistake is masking too tightly. If you paint exactly up to the outline of the object, the pixels immediately outside it are still part of that object — the edge blur, the colour fringing, the pixel or two where it fades into the background. The model treats those as context describing what belongs there, and faithfully rebuilds a faint ghost of the thing you were removing. Extend the mask a few pixels past the edge and the ghost disappears.

Cover the whole footprint, not just the object. Shadows are the usual oversight: erase a bin from a sunlit pavement and leave its shadow, and the result looks stranger than the original because the scene no longer makes sense. Reflections in water, glass and polished floors behave the same way. If it exists because the object exists, it goes in the mask.

Work one object at a time. A single mask covering several separate things scattered across the frame gives the model more to invent and less context to invent it from, and one bad region spoils an otherwise good pass. Since each run is charged, it is tempting to batch — but a second targeted run on a clean result is a better use of six credits than one sprawling attempt.

What it does well, and where it struggles

Inpainting is strongest when the background it has to rebuild is texture rather than structure. Sand, grass, foliage, plain walls, open sky, carpet, water, blurred bokeh: these are all things where any plausible continuation is correct, because no viewer knows what the exact arrangement should have been. Removing a stray person from a beach, a bin from a lawn, a power line from a sky, or a photobomber from a portrait background are the cases it handles almost invisibly.

It struggles where the background has structure that must line up. Brickwork, tiling, window frames, railings, stair treads, printed text, repeated patterns and architectural edges all have to continue in a specific way, and a model producing something merely plausible will produce something that is subtly wrong. A rebuilt line of tiles that fails to align with the tiles either side is more distracting than the object you removed.

Faces and hands are their own category and are best avoided as fill content entirely. If a person you are removing overlaps another person, the model will attempt to reconstruct the hidden part of the second person, and human vision is exceptionally well tuned to noticing when that goes wrong. In those cases, crop instead: losing part of the frame is nearly always less conspicuous than an invented limb.

When to erase, and when to crop instead

Object removal is a powerful tool and frequently the wrong one. Before running it, check whether cropping solves the problem. If the thing you want gone is near an edge of the frame — the usual case for stray bystanders, parked cars, signage and rubbish bins — a tighter crop removes it perfectly, instantly, for free, and with no possibility of an artefact. Cropping is lossless. Inpainting is a guess.

Reach for the eraser when the object is somewhere a crop cannot reach: in the middle of the frame, overlapping the subject, or positioned such that cropping it out would take the subject with it. Also reach for it when the composition depends on the current framing and you are not willing to trade any of it away.

There is a middle path worth knowing about too. If the goal is to obscure something rather than to make it vanish — a licence plate, a face, an address on an envelope, a name on a screenshot — blurring or redacting is the right answer, and it runs entirely in your browser for free. It is also the more honest choice, because a redaction is visibly a redaction, while an erased object silently changes what the photograph claims happened.

How the run works, and what it costs

Object removal is one of the OpusImg tools that cannot run locally. Inpainting requires a GPU, so when you press erase, your image and the mask you painted are uploaded, queued for a worker, processed, and the result is returned to you. The page marks this with a server-processing badge, and the same badge appears on every other tool that leaves your device, so the distinction is never hidden.

Each run costs 3 credits, charged when the job runs rather than when you like the outcome, because the GPU time is spent either way. That is the honest reason to spend an extra few seconds on the mask: a generous, complete mask on the first attempt is cheaper than a tight one plus a retry. Object removal is a Pro feature — you can open the tool and paint a mask on any account, but running the model needs a paid plan with credits available.

The output is always a PNG, named after your original with an -erased suffix. PNG is lossless, so the freshly generated region is not immediately handed a round of JPEG compression on top. If you need a smaller file for the web afterwards, compress the result as a separate step, where you can see exactly what the compression costs you.

Frequently asked questions

How do I remove an object from a photo?

Open the image, paint over whatever you want gone with the brush, and run the eraser. The model reconstructs the covered area from the surrounding scene, so the object is replaced by a plausible continuation of the background rather than by a blur or a patch of flat colour. Each run costs 3 credits and takes a few seconds on a GPU.

What is a magic eraser, and how does it work?

A magic eraser is an inpainting model. You supply an image and a mask marking the region to discard; the model then generates new pixels for that region, conditioned on everything outside it. It is not copying another part of the photo and pasting it in — it is predicting what the scene most likely looks like behind the object, based on patterns learned from a very large number of images.

How much does it cost to erase an object?

Three credits per run, and the run is charged whether or not you like the result — the GPU work happens either way. That makes the mask worth a few extra seconds of care. Object removal is a Pro feature; free accounts can open the tool and try the brush, but running the model requires a paid plan with credits available.

Can I remove people from a photo?

Yes, and it is one of the cases inpainting handles best when the background behind them is simple — a wall, sand, grass, open water, an empty street. It handles it worst when the person overlaps something structured that has to line up afterwards, such as a railing, a row of windows or a patterned floor. Removing a distant background figure is usually convincing; removing someone standing in front of a bookshelf usually is not.

Can I remove a watermark from an image?

Technically the tool can reconstruct the area under one, but removing a watermark from an image you do not own is copyright infringement, and it is not what this is for. The legitimate uses are erasing your own watermark from your own work, removing a timestamp burned in by a camera, and cleaning up a photograph you have the rights to.

How do I get the best result?

Cover the object completely, including its shadow and any reflection, and extend the mask a few pixels past its edges. A mask that stops at the outline leaves a faint halo of the original object behind, which is the single most common cause of a disappointing result. Beyond that, mask one object at a time rather than several scattered across the frame, and prefer a slightly generous mask to a tight one.

Why does the erased area look smudged or wrong?

Usually one of three things. The mask was too tight, leaving edge pixels of the original object for the model to build on. The masked region was very large, so there was little context to infer from. Or the background behind the object was structured — text, faces, brickwork, tiling — and the model produced something plausible in texture but wrong in detail. Re-running with a wider mask fixes the first; the other two are limits of the technique.

What image formats and sizes are supported?

JPG, PNG and WebP up to 25 MB. The result always comes back as a PNG, named after your original with an -erased suffix, because PNG is lossless and avoids adding a fresh round of compression artefacts on top of freshly generated pixels.

Is my image uploaded to a server?

Yes — unlike the browser-based tools on OpusImg, object removal needs a GPU, so the image and your mask are uploaded, processed, and returned. This is stated on the page with a server-processing badge. If a file must never leave your device, use the local tools instead: crop, blur and redact, resize and compress all run entirely in your browser.

Is this better than the clone stamp in Photoshop?

For most everyday removals, yes, and considerably faster — a clone stamp requires you to choose a source area and blend it by hand, while inpainting reconstructs the region in one pass. Manual retouching still wins where precision matters more than speed: matching a specific texture, working on a large area with structure in it, or fixing the parts of a generated result that came out wrong.