Resemblance is a style-family question first
Animators do not draw real faces; they simplify them into families, and every family has a bone-structure bias:
- Rubber-hose (1930s): round cheeks, soft chins, dot eyes. Real faces with wide cheeks and small chins land here.
- UPA mid-century flat: long faces, sharp angles, tiny noses, long limbs. Narrow faces with defined cheekbones map to this family.
- Saturday-morning (1980s to 1990s): big eyes, strong jaws, bold outlines. Even, symmetrical features with a defined chin fit.
- Anime: large iris-to-face ratio, small nose and mouth, gradient shading. Faces with big eyes and delicate lower thirds read here.
- Modern 3D: soft rounded volumes, expressive brows, button noses. Round-eyed faces with high cheekbones land in this family.
Your face maps to a family first and an archetype within it second, which is why one quiz answer can be right about the family and wrong about the character.
Why a photo beats a text quiz
A text quiz measures personality, and personality has no jawline. Face matching works on measurable geometry: the tool converts your selfie into an embedding, a list of numbers encoding eye spacing, nose bridge, cheek line, and chin width, then ranks the distance between your numbers and each archetype's reference. The result reproduces, because your geometry does not change with your mood; a quiz reshuffles every time you retake it. The same upload-a-selfie flow powers the photo lookalike tool behind the kpop idol matcher, and it runs the identical geometry-first logic.
Curious what you'd look like in this style? Preview it in about three minutes.
See a previewThe six cartoon archetypes
| Archetype | The tells | Family and styling cues |
|---|---|---|
| The Round-Faced Optimist | Round cheeks, wide smile, soft chin | Rubber-hose family; primary colors, simple bold shapes |
| The Lanky Deadpan | Long face, flat brows, dry half-smile | UPA family; muted palette, sharp collars |
| The Spiky-Haired Rebel | Angular jaw, strong brows, restless hair | Anime family; layered jackets, warm tones |
| The Big-Eyed Dreamer | Large eyes, delicate nose, soft chin | Anime and 80s families; pastels, gradient light |
| The Square-Jawed Lead | Even features, strong chin, symmetrical brows | Saturday-morning family; saturated colors, bold outlines |
| The Button-Nosed Trickster | Small nose, high cheekbones, arched brows | Modern 3D family; soft rims, expressive styling |
The references describe animation traditions, not licensed characters. The matcher maps your features to the archetype, and the styling cues are generic on purpose, because that is what makes a resemblance you can actually wear.
Shoot the selfie the matcher reads
- Face a window with flat daylight, no direct sun, no lamps mixed in.
- Phone at eye level. Low angles stretch the chin and push faces toward one family falsely.
- Straight-on, not angled. The families split on jaw and cheek shape, and a three-quarter turn hides exactly those.
- Neutral expression, hair off the face. A grin flattens the cheek read; bangs hide the brow.
- No filters, no glasses. Beauty modes sand down exactly the features being measured, and lens glare shifts the eye read.
- Shoot three frames across three days. The archetype that repeats is your stable match.
Turn the archetype into a look
Each archetype converts into styling you can actually run. Round-Faced Optimists suit flat primary-color palettes and simple silhouettes; Lanky Deadpans suit sharp collars and monochrome; Square-Jawed Leads carry saturated jackets and graphic tees; Big-Eyed Dreamers suit pastel layers and soft light. For the anime family specifically, the anime look alike guide covers the styling in depth, the which superhero do i look like matcher handles the caped branch of the Square-Jawed Lead, and the 3D family gets its own treatment in which pixar character do i look like.
Where the match goes wrong
- Beauty filters. They widen eyes and shrink noses, pushing nearly every face toward the Big-Eyed Dreamer.
- Cosplay makeup in the input. Line and shadow belong to the output look, not the input photo.
- Group shots and distant crops. One face per frame, close enough to fill a third of the image.
- One-frame verdicts. A weird angle produces a confident wrong answer; trust the repeat.
- Reading too much into it. The result is a resemblance, not a personality report; treat it as a styling direction.
Upload five selfies, pick a style, get results back in about three minutes.
Try the generatorThe AI route: a cartoon render of your own face
If the archetype made you want to see it, the AI route draws the look without touching anyone's trademark. MyPhotoAI turns 5 to 15 selfies into portraits in cartoon-family and character-inspired styles with generic names: 33 creative looks inside a catalog of 107 looks across 6 categories (42 headshots, 33 creative, 15 boudoir, 7 lifestyle, 5 dating, 5 events). It does not promise "you as" any specific copyrighted character; the honest framing is style-inspired, the archetype's drawing language applied to your face. Previews arrive in about 3 minutes, watermarked and free to check; HD downloads cost money, with Starter at $15 for 5 photos, Pro at $35 for 20, and Premium at $65 for 50. Only upload photos of yourself.
It will not reproduce a studio's exact character model, and it works best as the last step after the match: archetype first, render second. For a creature-flavored read instead of a cartoon one, the which monster are you guide runs the same face-first logic over a different menagerie.
