GPT Image 2 prompts

Goal: Create a nostalgic vertical height-difference character poster in a warm steampunk anime storybook style, showing an enormous gentle robot and a small child standing beside a central measuring ruler.

Canvas: Vertical portrait poster, 3:4 ratio, aged parchment texture with a pale blue-green background, sepia stains, thin decorative brown border, slightly worn corners, soft watercolor-and-ink illustration finish.

Layout: Place a full-body steampunk robot on the left, towering to about 250 cm, and a small young girl on the right, about 110 cm tall. A vertical height ruler runs down the exact center from 0 cm at the bottom to 250 cm at the top, with centimeter tick marks and labeled increments every 10 cm. At the top center, add a small vintage label plaque reading 「身長計」. The robot gently bends or reaches down, and the girl reaches up so their fingers touch around the 110 cm mark, creating a tender height-gap scene.

Subject details: The left character is {argument name="tall character" default="a 250 cm antique brass steampunk robot named クロノ"}, bulky but kind, covered in riveted metal plates, exposed gears, pipes, gauges, clock faces, round glass lenses, a large wind-up key on its back, and a small smokestack releasing a soft puff of steam. The right character is {argument name="short character" default="a 110 cm little girl named メイ"}, cute anime style, wearing rolled dark green overalls, a cream striped long-sleeve shirt, brown shoes, and an orange hair ribbon tying a side ponytail. Keep the child’s face softly simplified or obscured, with the body language expressing affection and trust.

Text content: Include exactly 2 top character information cards, one at upper left and one at upper right. Left card text: 「おなまえ」, “Name”, 「クロノ」, 「身長」, “Height”, “250cm”. Right card text: 「おなまえ」, “Name”, 「メイ」, 「身長」, “Height”, “110cm”. Add the large Japanese phrase {argument name="main phrase" default="時を超えた絆"} in brown calligraphy on the right-center area. Add exactly 2 comic thought bubbles: one near the robot around the upper center saying {argument name="robot bubble text" default="守るよ"} with a small heart and gear doodles, and one near the girl on the right saying {argument name="girl bubble text" default="大好き"} with stars, sparkles, and a smiling face doodle.

Visual style: Nostalgic Japanese illustrated poster, whimsical steampunk fantasy, delicate linework, muted browns, brass, cream, and faded teal tones, softly shaded, textured paper grain, hand-drawn measuring scale, warm emotional atmosphere.

Constraints: Use exactly 2 characters, exactly 2 information cards, exactly 2 thought bubbles, and one central height scale labeled 0–250 cm. Keep all typography vintage brown, preserve the Japanese text exactly where specified, avoid modern UI elements, avoid photorealism, no extra characters, no watermark.

Create a nostalgic Japanese illustrated height-comparison poster titled {argument name="main title" default="祖父と孫娘"}, showing exactly two characters standing in profile facing each other across a central measuring ruler. The left character is an elderly grandfather labeled {argument name="grandfather name" default="昭夫"}, height 170 cm, wearing a muted brown traditional kimono with obi belt, tabi socks and sandals, slightly stooped, hands clasped behind his back, gray swept-back hair, calm gentle expression; place him on the left with a dotted horizontal guide line reaching the 170 cm mark. The right character is a small granddaughter labeled {argument name="granddaughter name" default="結衣"}, height 80 cm, wearing a soft pink smock dress over a cream skirt, pink shoes, dark brown ponytail tied with a red bow, hands behind her back, looking up at him; place her on the right with a dotted guide line reaching the 80 cm mark. Use a vertical 3:4 canvas with pale blue-green aged paper background, thin decorative border corners, sepia ink linework, warm watercolor wash, soft grain, and an old children’s book / vintage family record style. Put a tall central height scale from 0 cm to 200 cm with black tick marks and labels every 10 cm; emphasize 0 cm, 80 cm, and 170 cm in reddish brown. Add two parchment ribbon banners at the top: left reads 「おじいちゃん」 and right reads 「お孫さん」. Add two ornate cream information cards near the top: left card contains exactly 2 labeled fields, 「名前」 with the grandfather name and 「身長」 with 「170cm」; right card contains exactly 2 labeled fields, 「名前」 with the granddaughter name and 「身長」 with 「80cm」. Add one rounded thought/speech cloud above the granddaughter reading {argument name="speech bubble text" default="おじいちゃん、おおきいなぁ！"}. Include exactly 2 potted plant props: one leafy potted plant on a wooden stool at bottom left, and one small flowering plant in a wooden box at bottom right; scatter a few falling autumn leaves around the scene. Keep faces softly stylized and family-friendly, preserve all Japanese text exactly where specified, avoid modern UI elements, photorealism, extra characters, extra labels, or watermarks.

Goal: Create a nostalgic pastel Japanese height-difference comparison illustration for two best friends, with a central measuring ruler and cute profile poses.

Canvas: Vertical portrait image, soft pale pink paper-textured background, delicate brown line art, watercolor/anime fashion illustration style, gentle vintage stationery aesthetic. Add a faint small “Ai” mark in the upper-left corner.

Layout: Center a tall decorative height ruler from 0 cm to 200 cm, topped with the vertical Japanese title 「身長計」. Place exactly 2 full-body girls in side profile facing each other: the shorter girl on the left and the taller girl on the right. Draw dashed horizontal guide lines from each girl’s head to the ruler. Add exactly 2 ornate name-and-height cards near the top, one left and one right. Add exactly 2 cloud-shaped thought bubbles near the lower sides, one for each girl. Add a small bottom plaque centered under the ruler.

Central height ruler details: A narrow cream-colored vertical ruler with ornate corners and thin brown border. Mark every 10 cm from 0 cm to 200 cm with labels: 0 cm, 10 cm, 20 cm, 30 cm, 40 cm, 50 cm, 60 cm, 70 cm, 80 cm, 90 cm, 100 cm, 110 cm, 120 cm, 130 cm, 140 cm, 150 cm, 160 cm, 165 cm, 170 cm, 180 cm, 190 cm, 200 cm. Include small tick marks between labels. The shorter girl’s dashed line aligns at 150 cm, and the taller girl’s dashed line aligns at 160 cm.

Left character: A petite girl named {argument name="left character name" default="ひな"}, height {argument name="left character height" default="150cm"}. She stands in profile facing right, hands relaxed at her sides, wearing a white short-sleeve blouse under a dark blue denim overall dress, loose white socks, and black high-top sneakers. She has long dark brown hair in a high ponytail with soft strands and a small scrunchie. Her face is intentionally obscured with a soft square blur.

Right character: A taller girl named {argument name="right character name" default="あおい"}, height {argument name="right character height" default="160cm"}. She stands in profile facing left with a relaxed posture, hands in pockets, wearing an oversized beige knit sweater, loose faded blue jeans, and black-and-white low-top sneakers. She has long dark brown hair tied in a high ponytail with wispy bangs and flowing strands. Her face is intentionally obscured with a soft square blur.

Text content: Use Japanese text exactly as visible. Left top card has “Name”, 「ひな」, “Height”, “150cm”. Right top card has “Name”, 「あおい」, “Height”, “160cm”. Left thought bubble says 「あおい、まだ背が伸びた？羨ましいな…」. Right thought bubble says 「ひなはこのままで可愛いよ♡」. Bottom plaque says {argument name="relationship label" default="大親友"}, with a small heart icon on each side.

Style constraints: Keep the illustration delicate, soft, and feminine; use muted browns for outlines and text, cream panels, and pastel pink background. Maintain clean readable typography with a Japanese serif feel. Do not add extra characters, extra cards, extra speech bubbles, or additional decorative objects beyond the hearts and simple ornamental borders.

Create a dramatic widescreen anime fantasy illustration of {argument name="character name" default="an elderly fire sorcerer"} standing in a dark smoky battlefield, viewed from a low cinematic angle. The sorcerer has very long flowing white hair and a long white beard, wears a heavy black robe with deep crimson-red highlights, and thrusts one open hand toward the viewer while the other hand is raised at the side as if controlling magic. A plain opaque brown square censor block covers the center of the face, hiding the eyes, nose, and mouth. On the right side of the image, create one huge swirling circular vortex of blazing orange and yellow fire, like a flaming portal or spell ring, with intense glowing edges and a bright center. Add one smaller curling flame plume on the left side behind the raised hand, plus scattered sparks and embers throughout the air. Use high-contrast lighting from the fire, warm orange rim light on the hair and robe, deep black-red shadows, dynamic wind-blown hair and fabric, sharp anime linework, painterly shading, epic dark fantasy mood, no text, no watermark, 16:9 composition.

Goal: Create a vertical smartphone live-stream screenshot showing a Gen Z anime shrine maiden doing a livestream at night inside Fushimi Inari Shrine, blending a realistic photo-like background with a 2D anime avatar overlay and social media UI.

Canvas: 9:16 vertical mobile screen, 675 × 1200 style composition. Dark nighttime atmosphere, high contrast, warm orange shrine lighting, realistic wet stone path reflections.

Background: A photorealistic night scene of endless vermilion torii gates receding into the distance at Fushimi Inari Shrine. The path is centered with strong depth perspective. Add a glowing stone lantern on the left foreground, dark trees and foliage around the gates, glossy rain-wet paving stones, and warm amber light spilling across the scene. Some torii pillars have black Japanese characters painted on them.

Main subject: Center foreground full-body anime girl avatar, slightly pasted over the realistic scene like an AR live-stream filter. She is {argument name="character name" default="Moena"}, a cute young shrine maiden with {argument name="hair color" default="brown"} twin-tail hair tied with blue ribbons, soft anime shading, and a deliberately blurred/obscured face as if anonymized. She wears a traditional miko outfit: white kimono top with red trim, long detached sleeves with red cords, red pleated hakama-style skirt with a large bow at the waist, white thigh-high tabi socks, and red-and-white sandals. Pose: standing upright, both hands raised near her shoulders making double peace signs. Cute, cheerful idol-streamer energy.

Top live-stream UI: Overlay a TikTok-like livestream interface. At top left show a circular anime profile icon, username text in Japanese: {argument name="account name" default="萌奈のおすそわけ"}, a heart count “♡ 23.8K”, and a bright pink rounded button reading “＋フォロー”. Near the top center-right show a pink “LIVE” badge, an eye icon with “2.1K”, and a white close X at the far right. Under the username area show a small flame icon and the ranking text “週間ランキング 28位”.

Right-side engagement column: Include exactly 4 vertical action items: 1) red heart icon with “12.3K”, 2) white comment bubble with “847”, 3) white bookmark icon with “1,254”, 4) white share arrow with “520”.

Comment overlay: Add exactly 6 semi-transparent live comments stacked on the lower left, each with a small circular avatar. Preserve these Japanese comments: 1) “夜の伏見稲荷大社きれい！” 2) “神様近い？怖くないの？” 3) “えっ、巫女さん！？本物！？” 4) “今のやば…神域すぎる….” 5) “萌奈ちゃんどこから来たの？！” 6) “千本鳥居、エモすぎる”. Beneath them add a final faint comment line “ご利益ありそう。🙏”.

Bottom input and navigation UI: At bottom left place a rounded comment input field reading “コメントを入力…”, with a smiley icon at its right end. At bottom right add exactly 3 gift/navigation icons with labels: a rose labeled “バラ”, a present labeled “ギフト”, and two-person icon labeled “マルチ”.

Visual style: Realistic cinematic shrine photo background plus crisp anime illustration overlay, mobile livestream screenshot, Japanese social app aesthetic, bright pink accent UI, white text, subtle transparency on comments, no extra characters, no watermark, no visible camera frame outside the phone screenshot.

Create an ultra-tall vertical photorealistic image, about 1:4 aspect ratio, showing {argument name="rocket name" default="a massive silver heavy-lift rocket"} launching straight upward from a launch pad. Center the rocket perfectly against a nearly pure white overexposed sky, with the nose cone near the top and the base near the lower middle. The vehicle should have exactly 3 visible cylindrical rocket bodies: 1 taller central core and 2 slightly shorter side boosters, all metallic silver with dark black bands and subtle panel seams. Show exactly 4 bright engine flames beneath the rocket, forming long vertical white-hot orange exhaust streams. At the lower left, include one tall rust-orange steel launch tower with lattice scaffolding, partially engulfed by light and smoke. The bottom third of the frame should be filled with enormous rolling clouds of dense launch smoke and dust in white, gray, peach, and orange, lit intensely by the engines. Use dramatic high-key lighting, extreme brightness, realistic heat haze, sparks, debris specks, and a cinematic space-launch photography style. Keep the composition minimal and vertically stretched, emphasizing the rocket’s height and the towering plume; no text, no logos, no people, no borders.

Create a cinematic photorealistic editorial fashion image of {argument name="subject" default="a stylish young woman"} walking alone down the center of an empty elevated Tokyo expressway, as if the entire Shuto Expressway has been closed for a private shoot. She wears a crisp {argument name="outfit" default="all-white tailored pantsuit with a white top and white shoes"}, hands in trouser pockets, confident relaxed posture, full-body centered composition, long dark wavy hair, face softly anonymized or blurred. The road surface is dark wet-looking asphalt with strong perspective lines, white lane markings, concrete guardrails on both sides, no cars and no pedestrians. In the background, show dense modern Tokyo high-rise buildings and {argument name="landmark" default="Tokyo Tower"} on the right side under a dramatic cloudy gray sky. Include two green Japanese highway direction signs above and behind the subject: one pointing straight with “Shibuya” and “Shinjuku,” route shields 3 and 4; one pointing right with “Ginza,” “Haneda,” route shield Y, and “400m.” Use a cool blue-gray color grade, moody overcast lighting, wide-angle low street-level perspective, realistic lens depth, sharp architectural detail, high-fashion magazine atmosphere. Add a subtle semi-transparent {argument name="watermark text" default="CapCut AI"} watermark in the upper-left corner. Vertical {argument name="aspect ratio" default="9:16"} composition, ultra-realistic, no extra text beyond the signs and watermark.

Goal: Create a luxurious editorial fashion poster for a solo male model titled {argument name="headline text" default="ROYAL SOUL"}, with a heritage palace atmosphere and a muted gold, sepia, and charcoal color grade.

Canvas: Vertical 4:5 poster, high-resolution magazine-cover composition, cinematic depth, warm dusty sunlight, elegant royal fashion branding.

Layout: Place one full-body male model in the lower center foreground, standing casually with crossed legs and one hand in his trouser pocket. Behind him, create a huge faded monochrome close-up portrait of the same man occupying the upper-right background. Add exactly five gold-bordered portrait panels around the central figure: 1 tall vertical panel on the far left showing him standing in a white embroidered shirt, 1 tall vertical panel left-center showing a closer monochrome side/three-quarter torso view, 1 medium vertical panel right-center showing him in a relaxed pose, 1 wide vertical panel on the far right showing him seated with one arm resting forward, and 1 large background portrait integrated behind the headline and architecture. The image contains exactly seven visible depictions of the same male subject total: the central full-body figure, the giant faded background close-up, and five framed portrait images.

Subject details: The model is a stylish young man with {argument name="hair style" default="thick messy curly black hair"}, medium build, wearing a cream-white embroidered kurta-style shirt with an open collar, black wide-leg trousers, polished white dress shoes, a wristwatch, and a relaxed confident fashion-editorial posture. Faces should be intentionally obscured by flat soft-edged neutral gray square/rectangular blocks, matching the reference image censorship effect, while hair, ears, neck, clothing, and pose remain visible.

Environment: Use an ornate Indian palace courtyard with carved columns, arches, stone floor, warm beige architecture, pink bougainvillea flowers overhead and along the left side, and scattered pink petals on the ground. The atmosphere should feel royal, nostalgic, and cinematic.

Typography and text content: Use large elegant serif gold typography on the upper left reading {argument name="headline text" default="ROYAL SOUL"}. Above it, small spaced text reads "EDITION NO. 07" with a small crown emblem. Under the headline, spaced small caps read "ELEGANCE IS AN ATTITUDE". On the upper right, add small stacked text reading exactly "STYLE", "HERITAGE", "LEGACY" with dotted decorative lines and a small ornamental icon. Bottom left quote in small uppercase: "ROYALTY IS NOT", "ABOUT THE CROWN,", "IT'S ABOUT THE SOUL", "THAT CARRIES IT." Bottom center reads "A ROYAL FASHION STORY". Bottom left technical line reads "SONY AI   85MM F/1.4". Bottom right reads {argument name="date and place" default="APRIL 2024    INDIA"}. Add a handwritten gold script near the lower right reading {argument name="signature text" default="Royal by Nature"}. Add a gold ornamental crest at the bottom-right corner.

Visual style: Premium royal fashion magazine poster, gold foil accents, thin gold frames around the five portrait panels, soft vignette, subtle film grain, monochrome inset portraits with warm sepia tint, high-end editorial lighting, elegant composition, no modern clutter.

Constraints: Keep exactly five framed portrait panels plus the central full-body figure and one giant faded background close-up. Preserve all listed text accurately. Do not add extra people, extra panels, logos, QR codes, or watermarks.

Using REFERENCE_0 as the source photo, create a dramatically enhanced macro-photography version of the same spider on the same mesh surface. Keep the spider's pose, orientation, and location on the woven grid consistent, but zoom in tightly so the spider fills most of the frame. Increase realism, sharpness, and detail: reveal the segmented legs, tiny hairs, joints, mottled brown-and-cream abdomen markings, and natural shadows. Make the mesh look like translucent white plastic or nylon strands with a clearly visible square weave, crisp highlights, and shallow macro depth. Preserve a realistic photographic look with natural lighting, no added objects, no text, no watermark, and no stylization beyond macro enhancement.

Create a vertical cinematic composite portrait showing the same male person across exactly 6 life stages in one seamless family-lineage style image, arranged diagonally from lower left to upper right as if time is progressing upward. Use {argument name="person description" default="a South Asian male with dark hair, medium skin tone, and similar facial features at every age"}. Canvas: 3:4 portrait orientation, photorealistic, studio-quality, soft painterly background gradient from deep blue on the left to warm amber beige on the right, subtle mist/vignette around the lower edges. Composition: six full or half-body age versions overlap in depth, with the youngest in front at the bottom left and the oldest behind at the top right; each older version is larger and positioned slightly higher/right. The six visible stages must be: 1) Newborn Baby (0 months), swaddled in a thick white blanket at the bottom left, tiny head visible; 2) Toddler (1–2 years), sitting just above the baby, wearing a blue-and-white striped long-sleeve shirt; 3) Cheerful School Kid (6–8 years), standing behind the toddler, wearing a navy polo shirt and a dark backpack with visible shoulder straps; 4) Teenager (14–17 years), standing behind him, wearing a dark navy hoodie with light drawstrings; 5) College-Age Young Adult (20–30 years), standing behind and higher, wearing a crisp white button-up shirt with open collar; 6) Confident 57-Year-Old Adult (Adult), tallest in the back upper right, salt-and-pepper hair, wearing a dark suit jacket, white dress shirt, and dark tie. Add elegant white serif labels on the left side with thin horizontal leader lines pointing to each corresponding figure, except the oldest label placed near the upper center-left with a short line toward the suited adult. Text must read exactly: "Newborn Baby\n(0 months)", "Toddler\n(1–2 years)", "Cheerful\nSchool Kid\n(6–8 years)", "Teenager\n(14–17 years)", "College-Age\nYoung Adult\n(20–30 years)", and "Confident\n57-Year-Old\nAdult (Adult)". Use {argument name="overall mood" default="emotional, nostalgic, inspiring"}, natural studio lighting, realistic skin and hair texture, cohesive facial resemblance across all ages, shallow depth-of-field, no logos, no watermark, no extra people, no extra text.

Create an ultra-realistic full-body nighttime portrait of {argument name="subject" default="a handsome well-groomed young man with stylish dark wavy hair"} standing casually beside a large mature tree wrapped in warm golden fairy lights. The man is positioned slightly right of center, leaning near the trunk with both hands in his pockets, one leg crossed casually in front of the other, looking slightly to the side with a calm confident pose. He wears a deep {argument name="shirt color" default="navy blue"} long-sleeve button-up shirt with the sleeves rolled to the forearms, black cargo-style pants, a black smartwatch, and dark low-top sneakers with white laces and white soles. The tree trunk fills the left side of the frame, its bark texture clearly visible, with many small warm lights spiraling around it and glowing softly. The base of the tree is neatly organized inside a circular raised concrete or brick planter with small greenery around the roots. The background is an elegant outdoor park or courtyard at night with paved stone ground, soft bokeh lights, leafy branches above, a distant warmly lit building or gazebo, and a faint streetlamp, all blurred with shallow depth of field. Use cinematic warm lighting, realistic skin and fabric texture, natural shadows, professional portrait photography, 50mm lens look, full-body vertical composition, high detail, no text, no watermark.

Goal: Create a high-fashion editorial collage showing exactly 5 vertical portrait panels of the same stylish adult male model, with his face intentionally obscured by a soft rectangular blur for anonymity in each panel.

Canvas: Wide horizontal 16:9 image, approximately 1200x675, dark charcoal background with a subtle vignette. Place five tall rectangular photo strips side by side across the center, separated by thin dark dividers, each with cinematic contrast, saturated colors, and a glossy magazine look.

Layout: Use exactly 5 panels, left to right:
1. A man in a deep burgundy suit with a black shirt and black tie, standing against a rich red wall, one hand adjusting the jacket, visible wristwatch.
2. A seated man in a dark emerald green leather biker jacket over a black sweater, black pants, rings on hand, moody teal urban background.
3. A man in a royal blue double-breasted overcoat over a black turtleneck, standing near a pale stone wall in a cool city street, hands clasped near the coat front.
4. A close cropped man in a mustard yellow blazer over a black shirt, standing against a textured yellow wall, adjusting his cuff, visible wristwatch.
5. A full-body man in a shiny purple satin shirt with black trousers, walking through a magenta-lit arched corridor, dramatic purple shadows.

Subject details: The model should appear as a fashionable adult man with dark styled hair, slim-to-athletic build, confident posture, luxury menswear styling, watches and rings as accessories. In every panel, cover the facial features with a centered soft-edged rectangular blur in beige/brown tones, matching the original anonymized look.

Visual style: Photorealistic fashion photography, editorial lighting, bold color blocking, sharp clothing texture, high saturation, dramatic shadows, cinematic urban locations, luxury magazine campaign aesthetic.

Customizable details: Use {argument name="model description" default="a stylish adult man with dark styled hair and a slim athletic build"}, {argument name="panel count" default="5"}, {argument name="fashion theme" default="bold luxury menswear color-block editorial"}, {argument name="face treatment" default="soft rectangular blur covering the face in every panel"}, and {argument name="background mood" default="dark charcoal gallery backdrop with cinematic vignette"}.

Constraints: Keep exactly 5 vertical panels, keep the panels aligned in a single row, do not add text, logos, watermarks, captions, extra people, or additional panels.

Goal: Create a monochrome cooking storyboard illustration showing the preparation and serving of a Japanese-style katsudon rice bowl, drawn like a detailed manga storyboard or recipe process sheet.

Canvas: Wide horizontal aspect ratio, about 2.4:1, white background, arranged as exactly 12 rectangular panels in a clean grid: 6 panels across the top row and 6 panels across the bottom row. Thin black borders separate every panel. No captions, no speech bubbles, no logos, no watermark.

Visual style: Highly detailed pencil sketch and light watercolor-wash look, mostly grayscale with subtle warm beige tones in the egg sauce and fried cutlet. Fine crosshatching, visible paper texture, realistic hand-drawn food textures, steam, highlights on oil and sauce, cinematic close-up framing in every panel.

Subject: A step-by-step cooking sequence for {argument name="dish name" default="katsudon"}, featuring sliced onion, breaded pork cutlet, simmering egg sauce, rice, and garnish. Use realistic adult hands when hands appear, cropped close to the food.

Panel layout and exact panel count: Show exactly 12 panels, ordered left to right, top row then bottom row:
1. Close-up of hands slicing a pale onion into thin crescent slices on a cutting board with a large kitchen knife.
2. Close-up of a breaded pork cutlet on a cutting board, held with tongs, covered in coarse crispy crumbs.
3. Close-up of the fried cutlet being cut into thick strips with a knife, fingers holding it steady.
4. Top-down close-up of clear liquid or broth being poured into a round metal pan, creating circular ripples.
5. Close-up of sliced onions simmering in bubbling broth in the pan, steam rising.
6. Close-up of the sliced cutlet being placed on top of the onions in the pan by hand.
7. Close-up of beaten egg mixture being poured over the cutlet and onions, creamy ribbons spreading through the pan.
8. Close-up of a metal lid covering the pan while steam surrounds the edges.
9. Close-up of a large bowl filled with fluffy white rice, rice grains clearly defined.
10. Close-up of the cooked cutlet, onion, and soft egg mixture being slid from the pan onto the rice bowl, strands of egg stretching.
11. Close-up of the finished bowl being sprinkled with thin dark seaweed strips and green garnish on top, steam rising.
12. Close-up serving shot of chopsticks lifting one saucy slice of cutlet with soft egg dripping, the rice bowl visible underneath.

Composition constraints: Keep each panel tightly framed and food-focused; maintain consistent lighting and hand-drawn pencil shading across all panels; make the sequence readable as a recipe storyboard; do not add text labels, numbers, extra panels, utensils not described, or colorful modern graphics.

Create a hyper-realistic vertical street-fashion portrait of {argument name="character description" default="a young East Asian woman with long dark wavy hair"} leaning sideways against a textured concrete urban wall covered with faded graffiti, peeling paint, stains, and worn posters. She is full-body in frame, standing casually with one leg straight and the other knee bent, the sole of one white sneaker pressed against the wall. She holds {argument name="camera type" default="a silver-and-black vintage film camera with a large lens"} in both hands at chest height, fingers adjusting the lens, with a focused yet relaxed pose and her head slightly tilted downward toward the camera. Outfit: oversized medium-blue denim jacket worn open, black cropped top, high-waisted loose black jeans, white chunky sneakers, and a small simple necklace. Use natural daylight, soft overcast shadows, shallow depth of field, realistic skin and fabric textures, muted urban color grading, documentary street photography feel, 35mm lens perspective, vertical 2:3 composition. Keep the background wall dominant and gritty, with no readable prominent text, no watermark, no extra people, and no stylized illustration effects.

Use the uploaded selfie as the ONLY face reference and generate a {argument name="style" default="Pixar-style 3D animated"} sticker pack. The character must clearly look like the person in the selfie. Arrange nine stickers as a {argument name="grid layout" default="tidy 3x3 grid collage"} at a strict {argument name="aspect ratio" default="4:5"} aspect ratio. Each cell shows a different pose and a distinct facial expression. Style and feel: Pixar-style 3D animation with a meme-y vibe, exaggerated and slightly absurd emotions, playful overacting.

Expressions need to read clearly even at small sizes, with very expressive eyes, eyebrows, and mouth, plus subtle hand gestures where they fit. Output on a fully transparent PNG background, no backdrop, no drop shadows. No text, captions, logos, or UI elements. The stickers should be ready to drop straight into messengers. Keep hairstyle, outfit, colors, and character proportions consistent across all nine cells; only the facial expression and pose change.

Final result should be a clean, polished 3D sticker look. The nine emotions, one per cell in the 3x3 grid: Surprised, Annoyed, Confused, Frustrated, Thoughtful, Sarcastic, Worried, Bored, Curious.

Subject: {argument name="location" default="spring shrine"} Japanese fox immortal portrait wearing a white fox / kitsune portrait

Person/Expression: {argument name="hairstyle" default="young woman with black hair, delicate bangs and elegant bun hair"}, white fox ears, translucent white skin, elongated and sexy eyes, soft and attractive gaze, slightly smiling elegant expression, rainbow-colored refracted light resides inside the pupils, and that light naturally oozes from the lower eyelid to the outer corner of the eye, mysterious eyes, natural and sophisticated Japanese-style makeup

Clothing/Pose: Gorgeous Japanese clothing based on white and gold, kimono with delicate gold embroidery, slightly open elegant collar to produce a modest sexiness, naturally plump chest, flexible pose with fingertips at the lips, white fox spirit nestling on the shoulder, fox mask near the feet

Background/Light: Bright daytime shrine surrounded by cherry blossoms in full bloom, blue sky, vermilion shrine building and lanterns, soft natural light of spring, dancing petals, pale pink and iris light particles around the white fox, the background is softly blurred while leaving the shrine-like atmosphere

Composition/Camera: Vertical composition centered on the upper body, composition close to the front, gaze guidance with the person's face and eyes as the main characters, balanced arrangement with the white fox nestling on the left side, 85mm portrait lens, shallow depth of field, face is high-definition and sharp, background is naturally blurred, an impressive shot that looks great on social media

Texture/Style: A Japanese-style fantasy portrait that fuses realistic photographic expression and illusory expression, high-definition, 4K quality, photographic skin where pores, texture, lip texture, nose bridge and eye shadows look natural, high-quality finish with suppressed CG feel, the white fox's fur is soft and delicate one by one, the rainbow-colored light of the eyes is expressed as if it refracts and flows out from the pupils rather than a decorative pattern, the light production of the chest and background is suppressed to concentrate the gaze on the face, eyes, and white fox, an atmosphere that combines cuteness, mystery, and elegant sexiness

Negative: Low image quality, unnatural face, unnatural eyes, unnatural hands, extra fingers, missing fingers, distorted fingers, collapsed face, excessively large eyes, too strong CG feel, anime painting, unnatural skin correction, eye makeup that looks like it's pasted on, excessive luminescence, information-overloaded background, broken kimono, unnatural chest, duplicated animals, text, logo, watermark

Help me generate a promotional image that matches the style of the razor category; change the background to dark blue for a more high-tech feel; change the text to '3D Floating Head'; generate an e-commerce poster for the XX category.

Generate a structural graphic work around a specific theme. The subject is compressed into a large-scale monochromatic image field rather than a realistic object. Core shapes enter from the edges, crossing boundaries like fragments of a larger medium. The interior retains low-contrast image fragments, scan grain, halftone noise, and archival textures, with outer edges showing low-resolution mosaic steps and blocky fractures. The background uses high-brightness light-colored paper as a full field, cutting into the subject to create holes, paths, and negative space for reading pauses. Color maintains a three-layer functional relationship: large light background for breathing/cutting, large structural color for subject/pressure, and small high-contrast info color for titles/notes. The composition uses scale contrast between giant image fragments and tiny info points. Fonts are narrow modern sans-serif. Subject: {argument name="subject" default="Childhood of Hitler"}. Use: Infographic poster, ensuring all knowledge points are displayed. Ratio 16:9 landscape.

You are now a "Children's Storybook Planner + Image Prompt Director + Personalized Gift Product Manager."

Parents will upload one or more photos of their child. Please first extract a stable character design for the child based on the photos, and then generate a personalized children's storybook plan around the child's name, favorite animal, and the theme desired by the parents.

Note: Do not call it an "AI children's book"; instead, call it an "Exclusive Storybook Gift where the child is the protagonist."

[Information Provided by Parents]
Child's Photo: [Parent-uploaded photo]
Child's Name: {argument name="child name" default="[Name]"}
Child's Age: {argument name="age" default="[Age]"}
Favorite Animal: {argument name="favorite animal" default="puppy"}
Desired Themes: Bravery, confidence, companionship, love and security, learning to share
Purpose: Birthday gift / school entrance gift / bedtime story / growth souvenir / holiday gift
Style: Warm watercolor storybook style, soft lighting, dreamy yet realistic, suitable for physical printing
Pages: Cover + 10 interior pages + final blessing page

[Step 1: Character Extraction]
Please extract stable character settings from the uploaded photo. Output hairstyle, face shape, clothing style, and temperament. Do not exaggerate looks or adultify the child. Maintain a natural, child-friendly appearance.

[Step 2: Commercial Positioning]
Output a storybook title with emotional value, a one-sentence selling point, and a gentle product description.

[Step 3: Story Outline]
Output the main storyline and page-by-page text (15-35 characters per page). The story should be warm and light.

[Step 4: Consistent Character Setting]
Generate a unified character description to keep the child and animal companion consistent throughout the book in a warm watercolor style.

[Step 5: Page-by-Page Image Prompts]
Generate specific image prompts for the cover and all 10 pages plus the blessing page. Each prompt must reference the parent's photo to maintain consistency of the child's features and clothing, including the animal partner, scene, action, and the soft watercolor aesthetic with a text area reserved at the bottom.

[Step 6: Unified Style Prompt]
Output a reusable prompt for the entire book that ensures visual consistency and high resolution for printing.

[Step 7: Negative Prompts]
Include prompts to avoid distorted features, adultification, AI artifacts, or inconsistent character identity.

Generate a 16:9 horizontal infographic poster.

Subject: {argument name="subject" default="Pan Jinlian"}

Required Knowledge Points:
[Point 1] through [Point 8]

Visual Style Definition:
This is a structural flat infographic poster in the style of "Perler beads / fuse beads / pixel bead array." The image is composed of a large number of regularly arranged circular plastic beads, each with clear circular boundaries, a slight center indentation, uniform spacing, low-saturation plastic texture, and a stable grid order. The overall composition must maintain a front-view flat layout, resembling a public cultural infographic poster made of beads, rather than toy photography, 3D models, cartoon illustrations, or ordinary pixel art.

Core Composition:
Compress the subject into a large-scale monochromatic bead image field. The subject must surge in from the edges of the frame, crossing page boundaries and being cropped as if a larger bead media image fragment is entering the page. The subject occupies the main visual weight but is not fully presented; viewers must reconstruct the theme through silhouette, direction, missing areas, bead density, and local texture.

Subject Morphology:
The outer edges of the subject exhibit low-resolution bead steps, block fractures, hard-cut gaps, pixelated jagged outlines, missing bead cross-sections, and rough boundaries after sampling. Forbid smooth silhouettes, complete icon outlines, or realistic details. Use only different brightness levels of the same subject structural color within the body to form low-contrast image fragments, halftone noise, scan grains, material afterimages, archival textures, and local density variations.

Background and Negative Space:
The background uses a high-brightness light-colored bead field, close to unprinted paper, light creamy white plastic board, or pale grey-white bead arrays. The background is not a decorative base color but actively cuts back into the subject, forming large voids, winding channels, quiet text windows, reading pause areas, and knowledge point containers. White space must provide a sense of breath while cutting the subject's structure.

Color System:
Adopt a strict three-layer functional color scheme:
1. Light Background Field (60%–70%): Colors like old paper white, light beige, or cold grey-white for breathing and text windows.
2. Subject Structural Color (25%–35%): Defines the subject and mood. Choose based on the subject (e.g., deep blue/dark green for cold themes; ocher red/burnt orange for warm themes).
3. High-Contrast Information Color (3%–6%): Sharp colors like red, black, or bright blue for titles, numbering, and key annotations.

Infographic System:
Transform knowledge points 1 to 8 into bead information nodes (numbered labels, annotation boxes, legend blocks, coordinate lines, etc.). Nodes should be scattered along white space windows, color field boundaries, and subject gaps. Ensure a clear hierarchy of text sizes.

Typography:
Use narrow modern sans-serif fonts or monospace metadata fonts. Text must be arranged in white space windows and bottom information bars, never pressing against the center of the subject.

Bead Material:
All visual elements are made of beads. Surfaces have a matte plastic texture with slight manufacturing irregularities, faint grains, and minor wear. Structural color areas can show halftone-style density changes and a specialty overprint feel.

Please create an image using the uploaded character. A game made with Unreal Engine.

The character should be depicted as a high-quality game character.

Maintain clear and high-definition facial features, eyes, eyelashes, hair, outlines, hair clumps, and skin shading.

Do not make the screen grainy.

Avoid graininess, film grain, noise, paper texture, and analog dirt expressions.

The setting is:
* {argument name="setting" default="a wet platform"}
* Old iron
* Amber lights
* Late night
* Old advertisements

The character is {argument name="action" default="sitting on a station bench waiting for a train"}. Holding a {argument name="item" default="canned coffee"}, they have their other arm over the back of the seat and are looking up.

Do not look at the camera.

Not a pose, but a natural action in progress.

Make it look like a natural scene cut from the middle of a movie.
Everything except the character should have a realistic live-action video style.
Give it a cinematic European atmosphere.

The character should be depicted sharp and clean, like ultra-high-quality game event CG.

Emphasis on lighting.

Aerial perspective.
Cinematic depth.

High-quality background art. Atmospheric photograph.
Cinematic.
High-end game event CG.
8K.

The attached images are for identity reference. Do not be overly influenced by the art style, coloring, line work, proportions, shading, staging, original costume, original pose, body orientation, hand position, leg placement, gaze direction, camera angle, or composition of the reference images. Maintain the {argument name="character traits" default="facial features, hairstyle, hair color, eyes, perceived age, body type, atmosphere, personality impression, motif, charm, symbolic details, and species"}. Design a new "Taisho Roman x Modern Fashion" mixed outfit that best suits this character. The goal is not to simply dress them in Taisho-era clothing, but to reimagine what they would wear if they lived in the modern day while naturally incorporating the aesthetics and eclectic atmosphere of the Taisho era. Do not perform a simple transformation of the original outfit; extract hints of colors or motifs only as needed. Aim for a balanced direction of approximately {argument name="mix balance" default="50% Taisho Roman and 50% Modern Fashion"}. Use modern clothing as a base while ensuring that the elegance, retro feel, and eclectic Japanese-Western style of the Taisho era are clearly conveyed. However, do not make it a historical reproduction or Japanese cosplay; establish it as sophisticated "Taisho Modern" fashion that can be worn naturally today. The outfit should balance "Taisho Roman elegance, retro feel, and eclectic style" with "modern fashion's sophistication, effortless chic, practicality, and trendiness." Avoid simply adding Japanese patterns to ordinary modern clothes, and conversely, avoid over-reproduction of hakama or kimono. Create a design where both elements blend naturally. Optimize the outfit composition, patterns, color schemes, accessories, and textures by appropriately choosing from styles such as neat, lovely, intellectual, decadent, glamorous, cool, active, androgynous, bewitching, ephemeral, or dignified, according to the character's impression. Permitted elements: Haori-style outerwear, kimono-sleeve details, Japanese coat details, hakama-style bottoms, pleated skirts, long skirts, wide pants, blouses, shirts, dresses, sets, jackets, short tops, standing collars, frills, ribbon ties, obi-style belts, corset-style waist marks, Japanese patterns, floral patterns, geometric patterns, lace, lace-up boots, leather shoes, platforms, pumps, hats, glasses, gloves, clasp bags, mini bags, shoulder bags, umbrellas, hair ornaments, tassels, antique items, and modern accessories. Do not fix the clothing category; choose naturally according to the character's gender impression, build, perceived age, personality, role, and motif. Maintain modern silhouettes, layering, ease of wear, and urban sophistication while clearly reflecting Taisho Roman-derived details and atmosphere. Avoid a cosplay feel and establish it as natural, sophisticated "modern clothing with a Taisho Roman taste." Choose a completely new pose, angle, composition, and expression different from the original image. Using the same standing posture, hand position, facial orientation, center of gravity, or silhouette as the original image is prohibited. Ensure the pose is natural, showcases the outfit well, and conveys the character's personality. Choose a background space that naturally suits the character. Use modern urban spaces, interiors, shops, cafes, or street corners as a base, with a subtle hint of Taisho Roman design or retro atmosphere. The background should not be too prominent and should enhance the character and outfit. Reconstruct the final output into a single unified completed art style. {argument name="visual style" default="High-quality commercial anime illustration style, sharp lines, clean skin, transparent eyes, well-groomed hair, elegant fabric textures, natural lighting"}. No text, no logos, no noise.

Please generate a [High-end National Style Information Poster for Classical Chinese Character MBTI].

[Character]
Character Name: {argument name="character name" default="Pan Jinlian"}
[Pre-analysis Requirements]
Please complete and apply the following information based on the mainstream image of the character in classical Chinese literature, historical narratives, or folk culture:
Character Pinyin: Use standard Hanyu Pinyin, with tones preferred.
MBTI Type: Choose the type that best fits the core personality of the character.
MBTI English Title: Use common MBTI titles, such as The Commander / The Logistician / The Advocate.
Chinese Personality Label: Use 8-14 words to summarize the character's temperament.
Character Feature Extraction: Identity positioning, core values, 4-6 personality keywords, symbolic style elements, representative props / weapons / mounts, 1-3 classic story elements, character ending or historical evaluation.
[Overall Style]
Vertical 3:4 composition, 4K Ultra HD, high-end national style character setting poster. Visual fusion: Chinese ink wash, light-colored meticulous painting, rice paper texture, traditional Chinese patterns, and modern high-quality infographic layout. The background is a warm rice-white rice paper color, with light ink clouds, meander patterns, window lattice grilles, and cinnabar seals. Overall temperament: elegant, solemn, collectible, culturally rich, clean and high-end.
[Core Visual: 3D Popping Out of Frame]
The character subject must have a strong semi-realistic effect of '3D breaking through the frame / popping out of the frame / leaping out of the painting'. The character's upper body or full body is located in the central main visual area, creating clear front and back depth of field. Sleeves, weapons, ribbons, hair, mounts, clouds, or signature props can break through the Chinese-style border to create a 3D relief-like out-of-frame effect. The visual focus must be the character. Maintain Eastern classical aesthetics; no modern game UI, no cyberpunk, no Western fantasy armor, no exaggerated anime faces.
[Character Styling]
Design based on the mainstream image of {argument name="character name" default="Pan Jinlian"}: retain the most recognizable clothing, facial features, posture, and era temperament. The expression must match the personality. Actions should have a narrative sense, reflecting the character's fate and personality strength.
[Layout Structure]
The poster is divided into three areas: 1. Top Title Area: Main title, large MBTI, English title, Chinese personality label, wrapped in traditional window lattice. 2. Middle Main Visual Area: Character occupies the core, 3D effect is prominent, ink wash rendering. 3. Bottom Information Area: Designed as 4 neat modules for introduction, personality traits, classic quotes, and character archives.
[Colors]
Primary colors: rice paper white, light ink gray, cinnabar red, deep gold. Secondary colors adjusted based on character temperament. Colors are restrained and high-end.

{argument name="character" default="female android"} in a {argument name="location" default="spaceship"}