Inpainting
Inpainting regenerates only a selected (masked) region of an image while leaving the rest untouched - useful for removing objects, fixing details or replacing part of a scene.
Inpainting is targeted image-to-image: you paint a mask over the area you want to change, write a prompt for what should go there, and the model regenerates only inside the mask. Everything outside the mask is preserved, so edits blend seamlessly into the original.
What you can do with it
- Remove unwanted objects (a passerby, a logo, a power line) by masking them and prompting for the background.
- Fix problem areas like distorted hands or faces by regenerating just that patch.
- Swap or add details - change clothing, add an object, alter an expression - without touching the rest.
How it works
The model treats the masked region as the only part to denoise, while using the surrounding pixels as context so the new content matches the lighting, perspective and style. A denoising strength setting controls how freely the masked area can change versus how closely it follows the original underneath.
Mask a distracting object in a product photo, then prompt for the surface behind it to remove it cleanly.
Example prompt
clean marble countertop, soft studio lighting, matching the surrounding surfaceTry it in the generator
Put inpainting to work right now - free daily generations, commercial license included.
Related terms
- OutpaintingOutpainting extends an image beyond its original edges, generating new content that continues the scene - used to widen a crop, change the aspect ratio, or build out a larger composition.
- Image-to-imageImage-to-image (img2img) is the AI workflow that transforms an existing picture according to your prompt, keeping some of the original structure instead of generating from scratch.
- DenoisingDenoising is the core operation of a diffusion model: at each step it predicts and removes a little noise, gradually turning a random field into a clear image.
- ControlNetControlNet is an add-on that conditions a diffusion model on a reference structure - such as a pose skeleton, edge map or depth map - so you control composition precisely, not just with words.