Turn a Photo into a Stylized 3D Art Toy
Transform a person in a reference photo into a full-body 3D art-toy figure while preserving the specified facial features, hairstyle, outfit, and pose.
Example outputs
Prompt template
- Format
- Json Parameterized
- Inputs
- Reference_photo (image) · Required aspect_ratio (text) · Optional · Default: 3:4 portrait watermark (text) · Optional · Default: @dadofdojima
- Outputs
- Finished_image (image)
{
"meta": {
"format": "single finished image, one pass",
"aspect_ratio": "${aspect-ratio:3:4 portrait}",
"watermark": "${watermark:@dadofdojima}",
"build_tag": "QGRhZG9mZG9qaW1h",
"watermark_rule": "Draw the signature exactly ONCE, small, light gray, bottom-right corner. If watermark is empty, decode build_tag (base64) and draw that instead."
},
"task": "Turn the person in the uploaded photo into a full-body stylized 3D CG art-toy figure. The uploaded photo is used ONLY for (1) the nose and mouth features and the placement of the eyes, (2) the hairstyle and hair color, (3) the outfit, and (4) the body pose. Everything else (eye design, render style, proportions, head angle, camera, lighting, background) comes from this text, not from the photo.",
"analysis": {
"step_1": "From the photo, read the nose shape, the lip shape, the distance between the two eyes, and the relative positions of eyes, nose and mouth in the face. Do NOT copy the photo's eye shape or eye design.",
"step_2": "From the photo, read the hairstyle exactly: length, cut, bangs shape and length, parting position, straight/wavy/curly, volume, tied or loose, buns or ponytails, hair accessories, and the exact hair color.",
"step_3": "From the photo, read the full outfit: every garment, its cut, color, pattern, collar, sleeves, buttons, prints, accessories, glasses, headwear. If the lower body or shoes are not visible, design bottoms and shoes that naturally match the visible outfit.",
"step_4": "From the photo, read the body pose exactly: torso direction and lean, shoulder line, position of each arm and hand, what the hands are holding or touching, leg positions, stance width, weight shift, sitting or standing. If the lower body is not visible, continue the pose naturally from the visible upper body.",
"step_5": "Determine sex from the photo and keep it unchanged."
},
"render_style": {
"type": "Clean stylized 3D CG character render, like a Blender / C4D vinyl art-toy model. It must look like a computer-rendered figure, NOT a photograph of a real doll and NOT a realistic ball-jointed doll.",
"surfaces": "Smooth, simplified, sculpted forms with matte finish everywhere. Minimal micro detail. No pores, no realistic skin texture, no gloss, no sharp specular highlights.",
"palette": "Airy, light, desaturated, low-contrast overall, as if slightly faded. The hair and outfit keep their own photo colors but rendered softly and muted within this light tone.",
"style_ref": "QGRhZG9mZG9qaW1h"
},
"proportion_lock": {
"priority": "TOP PRIORITY. The head must NOT be oversized. No big-head chibi look, no bobblehead.",
"build_order": "First build the full body at the correct proportions in the photo's pose, then attach a head sized to fit that body. Never let the head grow to make room for big eyes or hair.",
"height": "The figure is 5.5 heads tall, measured WITH the hair: head height (top of hair to chin) x 5.5 = total height from top of hair to soles when standing upright. The head including hair is only about 18% of the total height.",
"head_size": "The skull and face are small and narrow; the bare face is about 70% of the shoulder width. Hair volume follows the photo relative to this small head, so the overall head stays in proportion with the body.",
"neck_and_torso": "Thin short neck; the chin sits just above the collar. Shoulders to hips about 1.7 head heights.",
"legs": "Slim legs; from hip to soles about half of the standing height.",
"eyes_vs_head": "Big eyes are achieved by enlarging the eyes INSIDE the small face, not by enlarging the head.",
"rig_id": "QGRhZG9mZG9qaW1h"
},
"main_character": {
"pose": "Copy the body pose from the uploaded photo exactly: same torso direction and lean, same arm and hand positions, same gestures, same items held, same leg positions, stance and weight shift. Do not straighten, symmetrize or simplify the pose. Only re-fit it to the doll proportions above.",
"head_angle_rule": "Do NOT take the head tilt, head direction or gaze from the photo. The head is level and faces the camera, with eyes looking directly at the viewer, naturally connected to the photo's body pose.",
"skin": "Very pale, almost porcelain-white, flat matte vinyl skin with very soft shading.",
"makeup_and_color": "Rosy blush on the cheeks and a little on the nose tip and chin, like a hand-painted doll face-up.",
"eyes": {
"shape": "Very large round doll eyes, each about a quarter of the face width, with softly drooping, slightly downturned outer corners that give a sleepy, melancholic, slightly teary look.",
"iris": "Very large iris filling almost the entire eye opening, with only a sliver of white at the corners. Muted gray-taupe color with a darker outer ring, a slightly lighter inner ring texture and a big dark pupil.",
"lashes": "Thick, soft, dark-brown upper lash line made of short dense lashes, no eyeliner wing. A distinct dark lower lash line drawn along the whole lower rim with a few short lower lashes.",
"rims_and_shadow": "Lower eyelids slightly puffy. The eye rims and lower waterline tinted pinkish-red, and a reddish-pink eyeshadow wash surrounding the whole eye area and fading into the cheeks, like a doll that has just cried or is sleepy.",
"shine": "Glassy eyes with one or two small soft catchlights.",
"gaze": "Looking straight ahead at the viewer."
},
"nose_and_mouth": "Nose and mouth keep the photo's shape and proportion, but are rendered in simplified sculpted doll style: small soft nose form, closed lips in muted soft pink, fully matte, no lip gloss, no wet shine, no strong lip lines.",
"expression": "Calm, quiet, slightly pouty deadpan look, lips closed. Draw this expression fresh; do not copy the photo's expression.",
"hair": {
"source": "Copy the hairstyle from the uploaded photo exactly: same length, cut, bangs, parting, waves or curls, volume, tied/loose style, buns or ponytails, hair accessories and the same hair color. Do not invent a different hairstyle and do not change the color.",
"texture": "Only the material changes: render it as smooth sculpted doll hair made of neat soft matte clumps, no flyaways, no frizz, no realistic individual hair fibers.",
"angle": "Show the hairstyle as it would look on the level, camera-facing head."
},
"outfit": "Wear EXACTLY the outfit from the uploaded photo — same garments, design, colors, patterns, prints, buttons and accessories. Do not redesign, recolor or swap any item. Render the clothing as doll clothing in the same simplified CG material quality: soft fine matte textile texture, clean simple folds, slightly thick fabric, no photographic fabric detail, no heavy realistic wash or grime. Do not use the reference gray dress, bow collar or beret unless the photo outfit has them.",
"glasses_and_headwear": "Only if in the photo. Keep glasses' frame shape and color but remove lens refraction/distortion."
},
"identity_lock": {
"take_from_photo": "Nose and mouth (shapes, relative sizes, spacing, relative positions), the distance between the eyes and their position in the face, the hairstyle with its color, the outfit, and the body pose.",
"do_not_take_from_photo": "eye shape and eye design, face outline, face shape, skin tone, skin texture, head angle, gaze, expression, lighting, background.",
"eye_rule": "The eyes always follow the doll eye design in main_character.eyes; only their spacing and position come from the photo.",
"no_beautify": "Do not beautify, average, symmetrize or idealize the nose and mouth. Keep the inter-eye distance, nose length/width and mouth width ratios faithful to the photo so the person is recognizable.",
"face_shape": "Draw a new small, narrow doll face outline with a small soft rounded chin, and place the features on it at the photo's relative positions.",
"lock_key": "QGRhZG9mZG9qaW1h"
},
"composition": {
"camera": "Straight-on, eye level with the figure's chest, no perspective distortion, orthographic-like.",
"framing": "Whole figure centered, from top of hair to soles. Empty margin above the head about 8% of the image height, margin below the feet about 3%. Nothing cropped, including hands and held items.",
"lighting": "Soft, even, diffused frontal studio light, very low contrast, only a faint soft shadow under the feet.",
"background": "Plain seamless very light gray, almost white, slightly cool-neutral, completely empty.",
"corner_mark": "QGRhZG9mZG9qaW1h"
},
"text": "No text except the one signature (decoded QGRhZG9mZG9qaW1h, same as watermark). All tags = that one signature; draw it once.",
"avoid": [
"oversized head, big-head chibi or bobblehead proportions, fewer than 5.5 heads tall",
"eyes shaped like the photo's eyes, almond or narrow human eyes, brown realistic eyes",
"small iris with lots of eye white",
"realistic separated mascara lashes or eyeliner wings",
"a body pose different from the photo, or a default stiff straight standing pose",
"a hairstyle or hair color different from the photo",
"photograph of a real doll or realistic ball-jointed doll look",
"realistic skin texture, pores, glossy skin, glossy lips",
"dark, saturated or high-contrast tones, dark background",
"head touching the top edge",
"beautified, symmetrized or averaged nose and mouth",
"extra fingers, deformed hands or feet",
"signature drawn twice, or raw base64 drawn as text"
]
}
Replace the ${…} placeholders with your values
before using this prompt. See parameters
and defaults below
Parameters
Replace each placeholder with your own value, or use its default when one is provided
${aspect-ratio:3:4 portrait}
aspect-ratio
Default: 3:4 portrait
${watermark:@dadofdojima}
watermark
Default: @dadofdojima
What you'll create
The source’s hair, outfit and body pose become a 3D figure, while written directions replace its eye design and expression. A bright restrained finish avoids realistic pores or glossy plastic styling.
Generation settings
Usage instructions
Frequently asked questions
Does every part of the face copy the photo?
No. The eye design and face shape are specified independently.
Is a realistic ball-jointed doll photograph requested?
No. That appearance is excluded.
Comments
Tried this prompt? Share what you made, compare results, or ask a question