Remove Object from Photo
Brush over any unwanted object and AI will erase it, filling in the background naturally.
Drop your image here Tap to choose your image
JPG, PNG, WebP, GIF, BMP, TIFF • Max 20 MB
Paint over the object you want to remove, then click Erase Object.
Removing object with AI...
This may take up to 30 seconds depending on image size.
Error message
How to Remove Objects from Photos
Upload Photo
Drag and drop your image (JPG, PNG, WebP, GIF, BMP, or TIFF) into the tool above, or click to browse. Up to 20 MB.
Brush Over Object
Use the brush to paint over the unwanted object. Adjust the brush size as needed. The red overlay shows what will be erased. Click Erase Object when ready.
Download Clean Photo
Compare the original and cleaned versions side by side, then download your photo with the object removed. You can also go back to edit again.
What Can You Remove?
People in the Background
Remove strangers, passersby, or photo bombers from travel photos, landmarks, and scenic shots. The AI reconstructs the background behind them naturally.
Text & Watermarks
Erase overlaid text, date stamps, watermarks, and logos from photos. Brush over the text and the AI fills in the background that was hidden underneath.
Power Lines & Wires
Clean up skyline and architectural photos by removing distracting power lines, cables, and utility wires. The AI seamlessly fills in the sky or building behind them.
Clutter & Distractions
Remove trash cans, signs, vehicles, or any unwanted objects that distract from the main subject. Great for real estate photos, product shots, and social media content.
How AI Object Removal Works
This tool uses LaMa (Large Mask Inpainting), a state-of-the-art deep learning model designed specifically for image inpainting — the task of filling in missing or masked regions of an image with realistic content.
What makes LaMa different from older inpainting methods is its use of Fast Fourier Convolutions (FFCs). Traditional convolutional neural networks only look at small local neighborhoods of pixels at a time. FFCs give the model a global receptive field, meaning it can see the entire image at once. This allows LaMa to understand large-scale structures, repeating patterns, and distant context — producing fills that respect the overall scene geometry and texture.
When you brush over an object and click Erase, the tool:
- Takes your mask — the painted area that tells the AI what to remove.
- Analyzes the surrounding context — textures, colors, edges, and patterns around the masked region.
- Generates new pixels — fills the masked area with content that blends seamlessly into the surrounding image.
- Preserves everything else — pixels outside the mask remain untouched at full original quality.
Tips for Best Results
- Brush slightly beyond the edges — extend your brush a few pixels past the object boundaries. This ensures the AI has clean context to work with and avoids leaving behind faint outlines or shadow remnants.
- Use a larger brush for bigger objects — increase the brush size so you can cover the object in fewer strokes. Smooth, continuous coverage produces better results than many tiny dabs.
- Works best on natural backgrounds — grass, sky, water, walls, pavement, and other organic or textured surfaces produce the most convincing fills. Highly structured or unique backgrounds (like a specific book on a shelf) are harder to reconstruct.
- Remove multiple objects at once — paint over all unwanted objects before clicking Erase. The AI processes everything in one pass, which often produces more consistent results than removing objects one at a time.
- Use Undo if you overshoot — if you accidentally paint over the wrong area, click Undo to remove the last stroke, or Clear to start the mask over.