An avatar for social networks
Profile picture, banner, channel thumbnail: a drawn avatar is recognisable from afar and does not age. The 3D render keeps the most likeness, anime looks the most like an avatar. Generate in 1:1 for the profile picture.
Turn a photo into anime, cartoon or 3D with AI
A selfie redrawn three times by Flux Kontext, as Japanese anime, as 2D cartoon and as a 3D animated film character, the face staying recognisable. The full workflow, no account needed.
Interactive demo
Real workflow, real results. Pan and zoom freely.
Use this workflowThe real workflow: the same selfie enters three Image nodes set to Flux Kontext, each with its own style prompt, and you compare.
At a glance
- What the workflow produces
- 3 images in 3:4
- Canvas structure
- 8 nodes wired by 6 connections
- Models wired in
- Flux Kontext
- Cost of one full run
- about 18 credits, roughly $0.22
Seeing yourself as an anime hero, an animated series character or a 3D film figure: it is the format running on every network, and it rests on a single operation, image editing. This workflow takes your photo and redraws it three times with Flux Kontext, an editing model: it modifies the image you give it instead of inventing a new one, which is the only way to stay you.
All the value is in the prompt, and the demo shows it in full. Half of each text describes the style; the other half locks what makes you recognisable: face shape, hairstyle, glasses, beard, jacket, pose, framing, expression. Without that lock, the model draws a generic character in the requested style, pretty and anonymous. With it, anime flattens, cartoon exaggerates, 3D keeps the most likeness, and it is always the same person.
The demo selfie is generated, so as to use nobody's face. Yours enters through the Media node, and the three renders cost twenty-two cents, six credits per style. A front-facing, well-lit photo with no hand over the face gives the three best results.
How it works
- 1
Import your photo
A front-facing selfie, sharp, well lit, without sunglasses or a hand over the mouth. The model can only keep what it sees: what is hidden gets invented.
- 2
Adapt the lock to your face
In each prompt, the list "same hairstyle, same glasses, same jacket" describes the demo person. Replace it with yours: long hair, earrings, a cap, everything that is yours.
- 3
Run the three styles
"Generate all" redraws the three at once. Each node shows its price before the click, and the render replaces the built-in example.
- 4
Keep the one that looks like you
Compare the eyes and the jaw, that is where likeness is decided. The winning render downloads from the node or goes to an upscale for printing.
Variations and use cases
The same canvas, tuned for different needs: each variation is two or three lines of prompt away.
The whole family in the same style
One photo of each person, the same prompt for all, and the series comes out consistent: same line, same colours, same background. The consistent character guide explains why the fixed prompt matters more than the model.
A pet as a cartoon
The lock changes nature: coat colour, ear shape, collar, spot on the muzzle. The pet workflow goes further with staged scenes; here the style changes, not the scene.
A printed gift
The cartoon render of a couple or a child, enlarged and printed: run it first through the sharpen a photo workflow in creative mode, which rebuilds the strokes at poster size.
A character to animate next
The anime or 3D render is a first frame: cabled into a Video node, it becomes a character that moves. The transform a child's drawing workflow shows that sequence, image then video.
The same face in ten styles
Duplicate an Image node, change only the style description and keep the lock: black and white manga, Franco-Belgian comics, oil painting, pixel art. The lock is what holds the series together.
Common mistakes
Describing the style and forgetting the person
"Turn me into anime" with nothing else gives an anime character, not you. Half the prompt must name your features. It is the one mistake behind almost every disappointing result.
Starting from a blurry or three-quarter photo
The model only redraws what it sees. A blurry face, half turned or in shadow leaves gaps it fills with generic features. A front-facing, sharp photo before anything else.
Chaining two edits
Running the anime render through cartoon gives a cartoon of an anime, with the distortions of both. Each style starts from the original photo, which is why the three nodes are cabled to the same Media node.
Naming a studio or a film
The model interprets a studio name unpredictably and the result changes from one try to the next. Describe the line, the colours, the proportions: it is reproducible, and it is what the three demo prompts do.
Recommended models
Flux Kontext
Seedream 5 Pro
Topaz Bloom 2Frequently asked questions
Does it work on a group photo?
Yes, but the lock gets long: you have to name what distinguishes each person, or the model mixes two up. Three people at most to stay readable, and a photo where every face is sharp.
Can I ask for a specific style, a studio or a film?
Describe the style rather than naming it: "thin line, pastel colours, big eyes, watercolour background" works better than a studio name, which the model interprets freely. The three demo prompts are written that way, reuse them.
Why does the result not look like me?
Almost always because the lock does not describe your photo: it talks about a beard and round glasses when you have neither. Rewrite the list with your features, and check that the photo is front-facing and sharp.
How much does a full run cost?
Eighteen credits for the three styles, six per Flux Kontext render. The price shows on each node before you click. Re-running a single style only costs that style.
What render size?
Your photo's format, 3:4 in the demo. For a poster, enlarge the render afterwards with the Enhance node in creative mode, which rebuilds the drawing's strokes without blurring them.
Can I do it from my phone?
Yes, and the selfie is taken with it anyway. The canvas runs on mobile, the photo imports from the gallery into the Media node. The creating on mobile guide explains the gestures.
Can I use it as a public profile picture?
Yes, the render is yours. It is an image derived from your own photo by a licensed model, with no style copied from a named artist: nothing stands against public or commercial use.
Are my photos used to train models?
No. What you upload stays in your workspace and is handed to no training. The question matters more here than elsewhere, since it is your face.
Turn a photo into anime, cartoon or 3D with AI
Use this workflowFree account, no card required. The workflow opens pre-filled in your canvas.