AI Cartoon Generator: Create Cartoons from Text or Photos

Cartoon work rarely finishes in one pass. Here you can start from text, from a photo attached as an image reference, or from both at once, then keep editing the result without moving it into another tool.

You also get a choice of image models in the same workspace, so you can run one prompt through more than one model and compare how each handles line weight, shading, and likeness. From there the image can be edited, upscaled, resized, or used as the first frame of a video.

Open Create image, attach your photo with the Reference button, and describe the cartoon direction you want. Naming visual traits beats naming a show: “flat 2D cartoon, bold black outlines, simple geometric features, bright cel-shaded color” gives the model something concrete to follow. The same flow covers broader image editing once the cartoon exists.

Generate a few versions, compare them, and adjust the description if you want to change the pose, expression, line weight, or palette. Stylization shifts some facial detail by design, so aim for a result that keeps the subject recognizable rather than one that matches the photo detail for detail.

Paid plans include commercial usage rights for the content you generate, under our terms of use. That covers client work, merchandise, social campaigns, and branding.

Those rights apply to your generated content. They do not grant permission to use someone else's characters, logos, trademarks, or other protected material, and a cartoon built around an existing franchise stays that franchise's property regardless of how it was made. For commercial projects, build around original concepts or assets you already hold the rights to.

Anything you can put in visual terms. Flat 2D animation with bold outlines and cel shading, polished 3D characters with soft materials and cinematic light, retro rubber-hose cartoons with limited palettes and paper grain, comic art with ink linework and halftone shadows, loose editorial cartoons, and clean mascot designs built around a strong silhouette.

Describing the traits instead of naming a studio gives you more control and leaves you with an original concept. If what you actually want is manga and visual-novel conventions, large eyes and screentone shading and all, ask for anime instead.

No. Short, plain-language descriptions work, and prompt enhancement fills in what you leave out. Most people get a usable result from one sentence and tighten it from there.

Experience does help you ask for the right things. If you already think in line weight, palette, and composition, you will get where you are going faster, which is why the style notes on this page are written in those terms.

Cover subject, character design, cartoon medium, line style, color palette, pose or action, and background. “Cartoon dog” leaves all seven to the model. “Cartoon corgi in a red cape, rounded shapes, thick clean outlines, flat cel shading, bright yellow background, mid-stride running pose” leaves none of them.

Then iterate. Change one clause at a time so you can see what each change does. There are more worked examples in our prompt guide.