How is this different from the magic eraser?

The same engine with a different framing: this page is for removing one identifiable object, the eraser is for tidying an area. Use whichever matches what you are doing.

Can I remove a person?

Technically yes. Whether you should depends on whose photo it is — the terms of use cover this.

Does it work on large objects?

Less well. The more of the scene the model has to invent, the more likely the result looks wrong.

Is there a local alternative?

Cropping the object out of frame is lossless, instant and needs no upload. If the object is near an edge, try that first.

IA em nuvemBeta

Remove an object from a photo

Mark the thing you want gone and have the background behind it rebuilt.

Provedor de IA não configurado neste ambiente.

As ferramentas de IA em nuvem ficam desativadas até que um provedor de IA seja configurado no servidor. Nenhuma imagem é enviada e nenhum resultado é gerado.

Ver as ferramentas que já funcionam

This page has not been translated yet and is shown in English. It is excluded from search indexing for this language until it has been.

A parked car in a landscape, a stranger at the edge of a holiday photo, a plug socket in an interior shot — the distraction is usually well defined and easy to mark. Reconstructing what was behind it is the hard part, and the part a simple eraser cannot do.

This shares its engine with the magic eraser and differs only in intent: here you are removing one identifiable thing rather than tidying an area. Both send the image and a mask to the configured cloud provider, and both disclose that before anything is sent.

O que você ganha

Works from a rough mark
You do not need a precise selection. Cover the object generously and let the model decide where the boundary is.
Background is rebuilt, not blurred
The gap is regenerated from surrounding context instead of being smudged over.
Explicit upload disclosure
The provider is named on the page, above the button that sends the image.
Feeds the editor
Crop, resize and compress the cleaned image without re-uploading it anywhere.

Como funciona

1
Load the photo
Decoded in your browser; the mask you paint stays at the image's own resolution.
2
Mark the object
Brush over it. A generous mask usually beats a tight one.
3
Run
Image and mask are sent to the configured provider, and the result returns to the page.

Especificações

Beta. This tool is implemented and runs through the configured cloud AI provider. If no provider is configured on this deployment the page says so and processes nothing — it never falls back to a fabricated result. It stays out of search indexing until output quality is consistently good enough to be worth ranking.
Implemented via the AIProvider abstraction. Shares an engine with Magic Eraser.

Por que usar o KaiConvert

Sem upload nas ferramentas locais
Funciona offline depois que a página carrega
A IA em nuvem é sempre sinalizada antes de começar

Provedor de IA não configurado neste ambiente.

O que você ganha

Works from a rough mark

Background is rebuilt, not blurred

Explicit upload disclosure

Feeds the editor

Como funciona

Load the photo

Mark the object

Run