How AI Inpaint works
Use AI inpaint to remove objects or replace details in a photo. Brush over an area, describe the edit, then compare and download the result.
Use inpainting when one part of a photo needs a change. Mark the area you want to edit and describe the replacement or adjustment. Keep the selected region close to the object you want to change, then compare the generated result with your original photo.
Why Run AI Inpaint Here
Focused local edits
Upload, paint, generate, and save without a separate marketing page. Visitors and crawlers see the editor next to the explanation.
Region-limited generation
Paint the area to edit and describe the change. Preserving other areas is a goal, not a guarantee; compare the whole result with your original.
Prompt-driven local fills
Tell the workspace what should appear inside the region — a sunset sky, empty pavement, black sunglasses — instead of restyling the whole photograph.
Original vs edited comparison
Judge edges and texture with an interactive slider on the preview panel and again in the real example further down this page.
Browser-based — no installs
Run the AI inpaint online tool from a modern desktop or phone browser. Supported files include JPG, PNG, and WebP up to 20 MB.
Credit cost shown up front
Each run uses 6 credits, visible on the button before you generate. Most jobs finish in 30–90 seconds and download as PNG.
What You Can Fix with AI Inpaint
Local edits from a painted region plus a short prompt — useful for catalog shots, portraits, and social stills
Remove objects and stray clutter
Paint logos, wires, photobombers, or shelf mess and say what should replace them. The fill rebuilds pavement, fabric, or sky so the crop looks shot that way.
Add or swap small details
Place sunglasses, a hat, a sign, or a prop. Recolor a garment, refresh a dull sky, or restyle a tiny patch without booking another studio hour.
Clean product and portrait frames
Brush over the pixels that should change. Type a concrete instruction such as “replace the sky with a golden sunset” or “add black sunglasses.”
Prepare stills for ads and video
Fix one problem area, download the PNG, then continue in the image editor or send the still to the text-and-image video tool when you need motion.
How to Use AI Inpaint Online
Brush over the pixels that should change. Type a concrete instruction such as “replace the sky with a golden sunset” or “add black sunglasses.”
Upload your photo
Drop a JPG, PNG, or WebP up to 20 MB, or tap a built-in sample if you only want to try the workflow before using your own file.
Paint the region and describe the fill
Brush over the pixels that should change. Type a concrete instruction such as “replace the sky with a golden sunset” or “add black sunglasses.”
Generate and inspect
Click Inpaint (6 credits) and wait about 30–90 seconds. Drag the handle to compare original vs edited pixels along the edge of the region.
Download or paint again
Save the PNG, or clear the region and run a second pass. Many people fix clutter first, then add a small prop on a tighter brush.
Tips for Cleaner AI Inpaint Results
Paint slightly past the object you want gone, then ease the edge with the eraser so the region follows the real contour. A tight, honest selection gives the model structure to match; a sloppy blob that covers half a face will invent texture you did not ask for. If the first fill seams, shrink the region and generate again instead of describing the whole photograph in the prompt.
Prompts work best when they name the fill, not the problem. “Empty wooden tabletop, matching grain” beats “remove the cup.” “Soft overcast sky with no extra buildings” beats “fix the sky.” The model already knows which pixels you painted; it needs to know what those pixels should become. Mention material, color, and lighting when the patch has to sit next to a product or a face.
Start from a sharp, well-exposed file. Heavy compression, motion blur, or a tiny crop gives the network less truthful texture to copy outside the region. For text stamped on a photo, the dedicated text-removal tool is usually cleaner than a generic fill. For a full restyle of the subject, use the image editor or a stylizer rather than forcing this AI inpaint tool to redraw everything you did not paint.
A Real AI Inpaint Example
This portrait used the on-page sample, a painted region over the eyes, and a short sunglasses prompt. Drag the handle to inspect original vs edited pixels — the rest of the face should match the source.
Drag the handle to inspect original vs edited pixels
Original photo
Edited result
From a real generation on this site. Upload your file, paint a region, and describe the fill to get a similar local edit.
Related tools after you download
A local fill is often one step. These descriptive links stay on Pixwit if you need text gone, a broader edit, a transparent cutout, or more credits.
- Remove Text from Photo
When the problem is stamped type or a watermark-like caption, a dedicated text-removal pass is usually cleaner than a generic fill.
- AI Image Editor
Use a broader instruction when you want global lighting, style, or composition changes instead of a single painted patch.
- Background Remover
Cut the subject onto a transparent PNG after the local fix, then drop it on a catalog or slide.
- Pricing and credits
See how signup credits and paid packs map to the 6-credit cost on this page and the rest of the studio.
Background reading: Inpainting (Wikipedia)
Common questions
What is AI inpaint?
- AI inpainting edits a selected area by reconstructing it from your instruction and the surrounding image. Use a close-fitting mask for object removal, replacement or background repair, then compare the full result.
How does the region brush work?
- The brush marks “rewrite these pixels” (rose overlay on screen). The eraser restores “keep these pixels.” Only the painted region is sent as the edit target. Keep the selection tight around the object or sky you actually want changed.
What prompts work best with the AI inpaint tool?
- Say what should appear inside the region: “replace the sky with orange sunset clouds,” “empty sidewalk, matching concrete,” “add red running shoes.” Vague lines like “make it better” give the model nothing to lock onto.
How much does each run cost?
- Welcome credits depend on eligibility and signup country (0, 20 or 40). Non-subscribers have a combined limit of one video or image generation per UTC day; this does not guarantee enough credits or access to every model. Check the current quote before generating. The selected model, duration, resolution, number of outputs and reference inputs determine the quote. Subscriptions and credit packs are explained on the pricing page.
Which image formats are supported?
- Upload JPG, PNG, or WebP up to 20 MB. Results download as PNG. Very large painted regions on high-resolution files can take closer to two minutes.
Will pixels outside the painted region change?
- Paint the area to edit and describe the change. Preserving other areas is a goal, not a guarantee; compare the whole result with your original.
How long does a generation take?
- Most runs finish in 30–90 seconds. Large regions or high-resolution uploads may take up to about two minutes depending on queue load.
Is my photo stored on your servers?
- Images are uploaded to process your request. See the Privacy Policy for storage and handling.
Start with AI inpaint on your own photo
Scroll up to the form, paint the region, describe the fill, and generate. Signup credits show on the button.