No. It is a construction: an edge-preserving smoothing flattens texture, a short palette is cut from your photo, and ink lines come from a difference of two blurs. Nothing is downloaded and no style model guesses at your picture. Learned cartoon styles would need a model, and this page does not claim one.

Why do edges look outlined?

Because the ink lines are drawn wherever your photo changes sharply in brightness, which is what a cartoonist's outline follows too. Lower Line darkness for a softer, painted look, or set it to zero for flat colour alone.

Can I make it look like anime?

Not honestly with this method. Anime and 3D cartoon looks redraw faces and proportions, which takes a learned model. This mode keeps your photo's shapes exactly and only flattens and outlines them.

Why do faces sometimes go blotchy?

With few colour steps, a cheek can fall across a boundary between two palette colours. Raise Colour steps, or Smoothness so the cheek reads as one area before the palette is applied.

Is the downloaded file the same as the preview?

Yes. The smoothing and the palette are computed on the same working copy at every size, and the line width is a fraction of the picture, so the saved file at full size has the same colours and line weight as the preview.

Cartoonize Photo Online — Free Cartoon Filter

A rebuild: your photo, remade

This is a rebuild: the photograph is remade as flat areas of colour with ink outlines. It is built from your photo in four steps — an edge-preserving smoothing that flattens texture but stops at edges, a palette of 6 to 24 colours cut from the smoothed photo, every pixel snapped to that palette, and dark lines drawn where brightness changes sharply.

It is not AI, and it does not redraw faces. If you want lines without colour, the pencil sketch is the better door; for blocky retro art, pixel art; for softer painted regions, the painting effect.

The controls

Smoothness: How hard the edge-preserving filter flattens texture before the palette is cut.
Colour steps: How many flat colours, from 6 to 24, cut from your photo by median cut.
Line weight and darkness: How wide and how dark the ink lines are; darkness 0 leaves flat colour alone.
Saturation boost: Pushes colour before the palette is cut, for a brighter comic look.

Where it falls short

Small faces lose their features: an eye a few pixels wide becomes one patch of colour.
Busy textures become many tiny patches; raise Smoothness or lower Colour steps.
It cannot produce anime or 3D cartoon styles, which need a learned model.

The strip, the preview and the saved file run the same code on the same settings, with every size a fraction of the picture, so the file you download is the frame you approved.

Nothing is uploaded

There's no upload step on this page — your picture is decoded in the tab, drawn to a canvas, and gone the moment you close it. The filter runs in this tab, on your device, and the file is written here too.

Questions

Is this AI?: No. It is a construction: an edge-preserving smoothing flattens texture, a short palette is cut from your photo, and ink lines come from a difference of two blurs. Nothing is downloaded and no style model guesses at your picture. Learned cartoon styles would need a model, and this page does not claim one.
Why do edges look outlined?: Because the ink lines are drawn wherever your photo changes sharply in brightness, which is what a cartoonist's outline follows too. Lower Line darkness for a softer, painted look, or set it to zero for flat colour alone.
Can I make it look like anime?: Not honestly with this method. Anime and 3D cartoon looks redraw faces and proportions, which takes a learned model. This mode keeps your photo's shapes exactly and only flattens and outlines them.
Why do faces sometimes go blotchy?: With few colour steps, a cheek can fall across a boundary between two palette colours. Raise Colour steps, or Smoothness so the cheek reads as one area before the palette is applied.
Is the downloaded file the same as the preview?: Yes. The smoothing and the palette are computed on the same working copy at every size, and the line width is a fraction of the picture, so the saved file at full size has the same colours and line weight as the preview.

Related on this site

Rendered on your computer, not on a server.