GPT Image 2 prompts

Do a moe personification of a {argument name="creature" default="Deathstalker"} in {argument name="form" default="Arachne style"}

Create a white t-shirt product mockup (flat lay, front view, plain white background).

The t-shirt design uses black line art only. No color fill. No shading. White shirt throughout.

Illustration style: hand-drawn with a soft pencil or crayon. Lines are slightly wobbly and uneven, with inconsistent pressure — some strokes faint, some darker. Outlines do not always close perfectly. The overall impression is a casual, childlike sketch — imperfect but charming. Minimal detail. No clean vector lines. No digital precision.

The t-shirt has four printed zones:

TOP CENTER: Reproduce the festival logo from the first attached image exactly as-is — do not alter the font, layout, or graphic style in any way. Place it prominently at the top center of the shirt. Make it large enough to read clearly. Do NOT apply the hand-drawn or yuru-kyara style to this logo.

TOP RIGHT: A hand-drawn jagged speech bubble (spiky, explosive edges — like a shouting manga effect), slightly smaller than the character illustration. Inside the bubble, write "{argument name="dialogue" default="[Insert dialogue here]"}" in clumsy, uneven hand-lettered characters. The bubble and text should feel comic and energetic.

MIDDLE: Redraw the character from the attached image in a loose, casual, yuru-kyara style — simplified, childlike, charmingly imperfect. Use the attached image as the sole reference for pose, expression, and content. Draw the character smaller — occupying no more than 40% of the shirt's height. Leave generous white space around the figure. The speech bubble may overlap with the top portion of the character illustration. Do not shrink the character to avoid overlap.

BOTTOM: The artist name "{argument name="artist name" default="[Insert artist name here]"}" hand-lettered in uneven, shaky characters — as if written slowly and clumsily by hand with a pencil.

Keep all elements well-spaced and balanced. Clean, minimal, no background decoration.

[Core Requirements] A portrait of a person in a {argument name="clothing color" default="light pastel pink"} sleeveless top under the warm golden sunlight of late afternoon. Maintain realistic proportions.

[Person's Pose and Expression] [Person Reference Images 1-11] The subject is standing in {argument name="pose" default="profile"} with a dreamy and slightly sad gaze toward the camera. The person looks at the camera only with their eyes, with their head slightly lowered and turned. Hair naturally flutters around the face due to the wind blowing from behind.

[Hair and Makeup Details] A dark-haired bun style tied loosely at the back, but many strands of hair are naturally scattered over the face and neck. The makeup is transparent and clean, with a natural blush on the cheeks and lips in a moist, natural color. Fine beads of sweat are visible on the skin.

[Lighting and Direction] A warm and intense halo light enters through the hair from the center of the frame, wrapping the woman's hair and facial contours in a golden glow. Light reflects off the skin, giving it a sheen.

[Texture and Color Mood] A soft, dreamy atmosphere and graininess, as if shot with a film camera. Colors harmonize warm gold, green, and soft light blue. The fine texture of the skin and the cotton texture of the sleeveless clothing are felt.

[Film and Camera Lens Depth/Angle] Shallow depth of field makes the subject sharp while the background is softly blurred. Captured from the front in a medium close-up. A soft and warm depiction characteristic of old-fashioned film lenses.

[Background Elements (Optional)] Blurred green leaves are visible in the background.

[Signature] A small Threads logo (operated by Meta) is located in the bottom right, with "{argument name="signature" default="CHAse"}" written small like a signature in white cursive above it.

Design a realistic restaurant interior based on this menu: {argument name="menu" default="[menu]"}, matching the cuisine, price level, mood and target audience, cinematic interior photography.

Use the uploaded portrait only to preserve the person’s facial identity, hairstyle, hairline, skin tone, and natural presence. Create a realistic cinematic portrait under a {argument name="tree type" default="blooming cherry blossom tree"} at {argument name="time of day" default="sunset"}, warm golden backlight, soft silhouette, falling {argument name="petals" default="pink petals"}, the person gently holding a red rose near the face, quiet romantic atmosphere, shallow depth of field, natural skin texture, realistic body proportions, soft spring outfit, elegant composition, 3:4 vertical photo, high-quality DSLR look, no text, no watermark

Negative Prompt:
anime, cartoon, plastic skin, over-smoothed face, fake beauty filter, distorted face, wrong identity, bad hands, extra fingers, unnatural body, harsh lighting, low resolution, overexposed, messy background, logo, watermark, text

Antique portolan chart of {argument name="region" default="[REGION]"}, 15th-century style. Aged vellum, red/black/gold rhumb line network, hand-lettered Latin coast names, sea monsters, caravel ships, compass rose cartouche, gothic title scroll. 1:1.

Transform the subject of the photo into a {argument name="item" default="toy-like figure"}. However, the person is depicted as a slightly exaggerated 3D character. Soft sculptural forms, matte textures, simplified geometry, cinematic clarity, and high visual polish. Keep the character recognizable while stylizing. Behind the figure, place a branded box printed with the attached image (the one before characterization). Place a circular pedestal in front of the box and have the figure sitting on it. Represent it in an {argument name="environment" default="indoor desk environment"}.

Goal: Create a high-fashion character wardrobe concept board for a small child, titled around {argument name="section number" default="SECTION 42"}, with a calm museum-catalog / sci-fi archive aesthetic.

Canvas: Vertical 2:3 poster on a warm ivory paper background with subtle grain, soft studio lighting, muted shadows, and lots of negative space. Use thin pale taupe linework to form a honeycomb of exactly 7 outlined hexagons: 1 large central hexagon and 6 surrounding wardrobe hexagons.

Layout: Place the central hexagon in the middle with a small geometric diamond emblem above widely spaced text reading {argument name="central title" default="SECTION 42"}. Surround it with exactly 6 labeled outfit panels: top center, upper left, upper right, lower left, lower right, and bottom center. Each outfit panel contains one full-body child mannequin/model on the left or center and exactly 4 small vertical material/detail swatches on the right. Keep the child’s face softly obscured/blurred and anonymous in every panel.

Subject: The same young child, about 5–7 years old, petite, with messy dark curly hair tied loosely back, neutral stance, realistic proportions, editorial fashion styling. Skin tone medium-light to tan, but face intentionally blurred. Render as a polished photorealistic fashion lookbook, not cartoonish.

Six wardrobe panels and exact contents:
1. Top center panel label: {argument name="top outfit label" default="RITUAL MOTION"}. Child wears a cropped cream wrap shirt, layered belts and waist pouch, loose olive harem trousers, wrist wraps, and beige strappy platform sandals. The 4 swatches show cream fabric, olive fabric, sandals, and rolled pouch/bag.
2. Upper left panel label: DISCIPLINE TRAINING. Child wears a rust-orange athletic crop top, black loose cropped training pants, red wrist wraps, and orange sneakers. The 4 swatches show orange fabric, red wrap coil, black fabric, and orange sneaker detail.
3. Upper right panel label: FIELD MAINTENANCE. Child wears dark denim overalls over a charcoal shirt, work gloves, rolled cuffs, and rugged brown boots. The 4 swatches show denim, worn denim detail, black gloves, and brown boots.
4. Lower left panel label: DEPTH EXPLORATION. Child wears a bulky dark vintage diving suit or deep-sea exploration outfit with round helmet, tubes, heavy gloves, utility details, and black boots. The 4 swatches show a brass porthole/helmet detail, mechanical suit texture, underwater silhouette scene, and black boot detail.
5. Lower right panel label: CEREMONY PRESENCE. Child wears an elegant flowing ivory ceremonial dress with draped sleeves, soft folds, simple necklace, and pale shoes. The 4 swatches show ivory draped fabric, sandy/beige texture, pendant necklace, and ivory fabric folds.
6. Bottom center panel label: WEEKEND UNWIND. Child wears an oversized red hockey jersey with the number 42 and a winged-wheel style crest, denim shorts, striped socks, and plush animal slippers. The 4 swatches show red-and-white jersey stripes, denim shorts, patterned sock, and plush duck slipper.

Typography: Use thin uppercase letter-spaced sans serif text in muted gray-brown. Panel labels are small and minimal, split over two lines where needed. Central text is larger with wide tracking: SECTION on one line and 42 below it.

Visual style: Minimal luxury editorial, archival wardrobe taxonomy, soft beige palette with muted olive, rust, denim blue, ivory, charcoal, and red accents. Photorealistic garments with tactile fabric detail, clean cutout presentation, soft shadows beneath each child, precise thin hexagon borders.

Constraints: Use exactly 6 full-body child figures, exactly 7 hexagons total, and exactly 24 swatch/detail rectangles total, 4 per outfit panel. No extra characters, no watermark, no dense body text, no bright background, no logos except the fictional jersey crest.

Using REFERENCE_0 as the base photo, retouch it into a clean, professional lifestyle shot while preserving the bouquet, clear glass vase, white ribbon text “JUST FOR YOU,” and the exact arrangement with 4 visible peach roses. Brighten the entire scene with warm natural sunlight coming through the window, improve exposure and color, make the greenery and roses look fresh and vivid, and add a soft shallow-depth-of-field look to the outdoor view. Neaten the composition by centering the vase and bouquet more elegantly, removing distracting table items such as the 2 gray placemats and dark chair backs, and making the tablecloth look freshly ironed, smoother, cleaner, and softly lit. Keep the window, sheer curtains, and cozy indoor setting, but make the result look polished, airy, and photorealistic, as if shot for a home decor or flower delivery advertisement.

Goal: Create a cozy overhead breakfast flat-lay photo with handwritten Japanese diary-style annotations, as if made for a warm morning social media post. The scene should feel realistic, softly edited, and slightly nostalgic.

Canvas: Wide 16:9 landscape image, about 1024×576. Use a brown-and-white gingham tablecloth covering the full background, with soft morning light, gentle shadows, and a slightly warm film-like tone.

Main layout: Place exactly 2 main food/drink areas. On the left, a decorative red-and-white porcelain teacup filled with black coffee sits on a matching saucer. On the right, a large white square plate holds the breakfast foods. Leave darkened negative space in the upper left for the title handwriting.

Food and object count: Include exactly 7 visible breakfast items/props on the plate area: 1 stack of toasted bread slices, 1 partly peeled ripe banana, 1 small mound of granola, 1 clear glass bowl of pale yellow apple compote with 1 dark prune on top, 1 brown egg, 1 small wooden egg cup or wooden coaster behind the egg, and 1 white square plate. The toast should be the visual center of the plate, with golden toasted surfaces and thick white bread edges. The banana should be at the upper left of the plate. The egg should be at the upper right. The glass bowl should be at the lower right. The granola should sit between the toast and bowl.

Handwritten text and annotation count: Add exactly 9 handwritten annotation groups in white/black ink, with arrows, circles, small hearts, sparkles, and doodles. Preserve the Japanese text as decorative handwriting:
1. Large title at top left: 「わたしの整う朝ごはん。」
2. Subtitle under title: 「シンプルだけど、ちゃんと満たされる。」
3. Circular note near coffee: 「今日も はじまりは 深い一杯から。」
4. Small note below coffee: 「ほっと息をつく、大切な時間。」
5. Torn-paper sticky note at bottom left with pink tape: 「朝の光を浴びて、心もからだも ゆっくり目覚めていく。」
6. Note above banana with arrow: 「完熟バナナで やさしい甘みを チャージ。」
7. Bubble note beside egg: 「たんぱく質で 1日を すっきり スタート！」
8. Note near toast with arrow: 「サクッと焼けた トーストが 一番のごちそう。」
9. Lower-right notes around granola, compote, and overall feeling: 「ザクザクのグラノーラで 食感と栄養をプラス。」, 「りんごとプルーンで 自然な甘さを からだにやさしく。」, and 「ちゃんと食べると、心までふんわり軽くなる。」

Visual style: Photorealistic breakfast photography mixed with casual hand-drawn overlay graphics. Use white ink for most handwriting, black ink for food labels on the plate, dotted speech bubbles, curved arrows, tiny stars, small hearts, and playful underline marks. The typography should look hand-lettered, not digital. Make the annotations readable but naturally imperfect.

Customizable elements: The title should read {argument name="headline text" default="わたしの整う朝ごはん。"}; the drink should be {argument name="morning drink" default="black coffee in a red porcelain teacup"}; the main breakfast should feature {argument name="main breakfast food" default="thick golden toast"}; the fruit should be {argument name="fruit item" default="ripe banana and apple compote with prune"}; the overall mood should be {argument name="mood" default="gentle, cozy, mindful morning"}.

Constraints: Keep the composition clean and appetizing, do not add extra foods beyond the counted items, do not add people or hands, avoid logos or watermarks, and keep the Japanese handwriting integrated into the image like a real edited breakfast journal photo.

Using REFERENCE_0 as the character base, keep the anime girl, pose, outfit, hairstyle, bag, cropped vertical composition, and soft illustrated rendering essentially unchanged. Replace only the background with a realistic modern multiplex cinema lobby: dark walls, warm ceiling cove lighting, glossy tiled floor, movie-poster displays, a bright red neon sign reading {argument name="cinema sign text" default="CINEMAS"} on the upper left, a digital showtime board on the upper right with exactly 5 time rows, a concession counter labeled {argument name="concession sign text" default="FOOD & DRINKS"}, and a freestanding wayfinding sign in the left foreground reading {argument name="wayfinding sign text" default="IMAX"} with a right arrow. Preserve the character’s anime style while making the lobby more cinematic, spacious, and photorealistic than the reference background. Do not add extra characters, logos, or legible movie titles beyond the specified signage.

Goal: Create a dramatic anime-style financial loss illustration of a fantasy elf noblewoman investor panicking at her desk after a stock portfolio crash.

Canvas: Square 1:1 composition, high-detail digital anime painting, cinematic lighting, dark office/trading room atmosphere, slightly cluttered foreground desk, no watermark.

Main subject: Center a young adult elf princess or queen with long flowing blonde hair, pointed ears, a gold spiked crown, white gloves, a red royal dress with ornate gold armor-like trim, a white cape, and a blue-purple gemstone pendant. She is seated behind a desk, gripping both sides of her head in despair. Cover the entire face area with a plain opaque beige square censor block, positioned centrally over the face, with no facial features visible.

Background and layout: On the left, place one large dark stock-market monitor showing Japanese loss information in red, including the headline {argument name="portfolio value label" default="持ち株評価額"}, a huge red negative value {argument name="main loss amount" default="-2,456,789円"}, and a smaller line {argument name="daily change text" default="前日比 -18.73%"}. Below it, include a red jagged downward line chart with y-axis numbers around 1,600 to 2,600 and time labels from 09:00 through 13:00, ending in a large downward red arrow.

Foreground desk items: Include exactly 8 visible paper-related elements: 1 printed loss report on the left labeled {argument name="loss report title" default="損益報告書"} with red amount “-2,456,789円”; 1 central portfolio list sheet labeled {argument name="holdings list title" default="保有銘柄一覧"} showing exactly 4 red percentage losses with downward arrows: “-15.20%”, “-21.45%”, “-17.88%”, and “-20.31%”; 1 loose paper at the front-left edge; 1 flying paper in the lower-left foreground; 1 floating document behind the desk on the left; 1 floating document on the right labeled “含み損” with a red downward zigzag arrow; and exactly 2 crumpled paper balls on the desk, one near the center-right and one near the lower-right. Also include one white teacup with gold decoration at the far right edge of the desk.

Emotional details: Add exactly 2 white spiral stress scribbles above the character’s head, one to the left and one to the right. Use scattered paper, downward red arrows, and tense body language to communicate panic, regret, and financial despair.

Visual style: Polished GPT-image-2 style anime illustration, sharp linework, rich highlights on gold ornaments, soft glowing rim light on hair, detailed fabric and armor, dark cool background contrasted with vivid red loss numbers. Keep all visible Japanese text legible and integrated into the scene.

Goal: Create a minimalist luxury brand concept presentation cover slide for {argument name="brand name" default="BRAND"}, using a split editorial layout with sparse typography and a cinematic architectural photo.

Canvas: 16:9 widescreen slide, clean white background on the left half and a full-height photographic image on the right half. Use a refined monochrome palette: white, charcoal, deep navy-black, cool gray, and soft daylight.

Layout: Divide the slide vertically at the center. The left 50% is a white design area with generous negative space; the right 50% is a photo panel. On the far left, place one thin vertical line near the margin, broken into two segments with the vertical word “BRAND” centered between them. Near the top-left area, set the large spaced-out headline “CONCEPT”. In the middle-left area, place a small abstract geometric logo made from exactly 2 dark navy square/rectangular blocks, followed by the large spaced-out title “KEY MESSAGE”. Under the logo/title row, add one short horizontal rule, then exactly 2 small supporting text lines: “Short statement line” and “Supporting statement line”.

Text content: Use {argument name="main headline" default="CONCEPT"} as the top headline, {argument name="key message text" default="KEY MESSAGE"} as the central title, and the two supporting lines {argument name="supporting copy" default="Short statement line / Supporting statement line"}. All typography should be uppercase or title-like where shown, with wide letter spacing, thin modern sans-serif font, dark navy-black color, and precise alignment.

Photo panel: The right half should show a quiet, high-end architectural interior with concrete walls, tall floor-to-ceiling windows, soft natural light, and strong diagonal shadows. Include exactly 1 human figure: {argument name="photo subject" default="a woman in a long black coat, seen in side profile, standing calmly and looking out the window"}. The figure should be positioned slightly right of center in the photo panel, elegant and contemplative, with a low bun hairstyle and minimal detail.

Visual style: Premium Japanese/Scandinavian editorial design, calm, spacious, museum-like, understated corporate branding, cinematic realism on the photo side, sharp typography on the white side.

Constraints: Keep the composition uncluttered. Use exactly 1 vertical brand label, exactly 1 headline, exactly 1 abstract logo, exactly 1 central title, exactly 1 horizontal rule, exactly 2 supporting text lines, and exactly 1 person in the photo. No extra icons, no mockup borders, no watermark.

Create a square decorative anime photo-card portrait with a glossy pastel frame and handwritten doodles. Center the composition on a cheerful decora-fashion anime girl posing close to the camera with both hands raised open near her face, palms forward, fingers spread, wearing colorful rings and bracelets. She has turquoise twin-tail or layered hair with pink and blue streaks, cat-ear-like hair accessories, many star and heart hair clips, and an extremely busy rainbow decora outfit: striped sleeves, layered neon accessories, beaded necklaces, heart charms, badges, plush-like pins, frilly rainbow skirt, hot pink vest, and shiny candy-colored textures. Obscure the face with a centered soft-edged rectangular blur/mosaic block in muted blue-gray to pink gradient, as if privacy-censored, while keeping the hair and hands visible. Use a vibrant kawaii Japanese street-fashion style, high saturation, glossy anime rendering, detailed accessories, sparkles, bokeh lights in the background, playful and cute mood.

Canvas and layout: square 1:1 image, the portrait sits inside a thick rounded rectangular Polaroid-style border. The border is a pastel gradient from peach pink to cream to light cyan, with slightly worn paper texture and rounded corners. The inner photo window fills most of the card, with a larger bottom margin reserved for large handwriting.

Frame doodles and text: include exactly 28 visible decorative doodle/text elements around the border: 1 handwritten date at the top reading {argument name="date text" default="5/4 23"}; 1 large handwritten bottom title reading {argument name="bottom title" default="DRAGON"} in hot pink marker; 1 small word on the right reading “FUN!” in blue; 1 small oval speech bubble near the bottom right reading {argument name="speech bubble text" default="Taiyaki"}; 1 small phrase along the bottom reading “COLOR MY WORLD” in purple; 1 vertical left-bottom word reading “LOVE” in blue; 1 tiny bottom-left word reading “coo” in blue; 1 green fish-shaped taiyaki doodle at bottom right; 2 smiley faces; 6 hearts of varying sizes; 6 stars; 2 arrows; 2 loops/swirls; 2 tiny flower or clover doodles; and 1 underline swoosh beneath the title. Make the doodles look hand-drawn with pink, blue, green, yellow, purple, and orange markers.

Subject details: the character’s outfit theme should be {argument name="fashion theme" default="rainbow decora kei with dragon-themed badges"}; add a visible round badge on the chest area that says “DRAGON” or suggests a dragon mascot. Keep the character cropped from mid-thigh/waist upward, seated or leaning forward, with the hands overlapping the border area slightly for a lively sticker-photo feel.

Constraints: no photorealism, no minimalist styling, no clean corporate typography, no extra people, no watermark. Preserve the face-censor rectangle and the dense sticker-photo scrapbook aesthetic.

Create a highly saturated dark-pop anime fantasy illustration of {argument name="character name" default="a Cheshire Cat inspired cat girl"} perched sideways on a thick tree branch in an enchanted forest at sunset. The character has an oversized mischievous feline head with purple and orange fur markings, large pointed ears with pink-blue inner ears, one huge glowing red-orange eye open, the other eye winking, long white whiskers, and a very wide toothy grin with sharp triangular teeth. Her body is humanoid but petite, wearing a gothic black puff-sleeve dress with white ruffled trim, a small black neck bow, lace details, and glossy black platform lace-up boots. Add exactly 5 prominent outfit elements: black puff-sleeve dress, white ruffled cuffs, black neck bow, striped arm sleeve, glossy platform boots. The visible legs wear bold horizontal striped thigh-high stockings in alternating pale blue, navy, cream, and warm peach tones. One arm has a vivid striped sleeve in red, pink, cyan, orange, and yellow, ending in a fluffy white cat paw glove with a ruffled cuff and small gold button. Include one huge curling fluffy tail sweeping across the lower foreground, colored with alternating purple and orange bands. Composition: vertical portrait, dynamic low-angle view, the character leaning forward playfully from the branch with legs dangling toward the viewer; one paw rests near the boots, the enormous tail frames the bottom. Background: tangled dark tree branches, glowing golden sun behind the head creating strong rim light, and stained-glass-like multicolored foliage in teal, magenta, yellow, orange, and violet. Visual style: {argument name="art style" default="glossy niji anime dark pop illustration"}, ultra-detailed linework, dramatic backlighting, neon color grading, high contrast shadows, luminous highlights, painterly texture, whimsical but slightly sinister Alice-in-Wonderland mood. Use {argument name="color palette" default="electric purple, orange, cyan, gold, magenta, and black"}. No text, no watermark, no frame, no extra characters.

Create a romantic vertical anime fantasy illustration shown as a smartphone screenshot preview with black UI bars at the top and bottom. The phone status bar at the top shows the time “23:06” on the left, signal bars, “4G”, and a battery icon on the right, plus a small white X close button near the upper left of the image area; the bottom black bar shows a white thumbs-up icon, a thin divider line, and a white thumbs-down icon. Inside the screenshot, depict an intimate birthday and wedding-anniversary celebration scene at night: a tall, broad-shouldered beastman husband with {argument name="male character species" default="lion-like cat ears, long blond mane, and large clawed hands"} embraces a petite young wife with {argument name="female character look" default="long wavy platinum-blonde hair, fair skin, delicate purple flower hair ornaments, and violet jewelry"}. Their faces are intentionally covered by two plain opaque censor rectangles in muted brown tones, one larger square over the man’s face and one smaller vertical rectangle over the woman’s face, while the rest of their bodies remain visible. The man wears an ornate black-and-gold embroidered formal jacket over a cream vest and cravat, with a dark choker-like collar and decorative chains; his arm wraps protectively around her waist. The woman wears a flowing lavender off-shoulder fantasy gown with sheer sleeves, floral lace, sparkling embroidery, and a soft glowing skirt; her hand rests gently on his chest. Set them on an elegant sofa in a candlelit room filled with warm golden bokeh, a dark window showing a deep blue night outside, flowers, lanterns, and wrapped gifts. In the foreground include exactly 5 prominent celebration elements: 1 round white cake decorated with pink and purple roses and candles, 3 lit candles on or near the cake, 1 gold crown-like cake topper, 2 wrapped gift boxes with lavender bows on the lower right, and scattered pink-purple blossoms. Use a luxurious shoujo fantasy anime style, highly detailed hair strands, soft rim light, glowing candles, romantic purple and gold color palette, shallow depth of field, sparkles, painterly rendering, and an intimate affectionate pose. No readable text inside the illustration except the smartphone UI indicators.

Using the provided reference image as the base for the same group selfie, regenerate it as a new nighttime nightclub selfie with the same three women and the same face-anonymized look. Change the composition so the woman on the left is closest to the camera with her arm extended taking the selfie, the center woman stands slightly behind in a fitted white ruched mini dress, and the woman on the right leans in while making a peace sign. Update the outfits to exactly 3 visible looks: left woman in a black camisole crop top with denim shorts and a chain-strap bag, center woman in a white mini dress, right woman in a white lace-up camisole crop top with denim shorts. Replace the background with a larger crowded club scene full of people, dark atmosphere, magenta/blue/purple stage lighting, laser beams, and a distant DJ/stage area. Keep the candid flash-photo aesthetic, wide-angle selfie distortion, glossy skin highlights, and realistic low-light club photography. Maintain solid rectangular face blocks over all 3 women’s faces, matching the anonymized style of the reference. Do not add text, logos, or extra main subjects.

Create a luminous fantasy anime illustration of a tender spiritual encounter between a radiant angel and a kneeling nun in a blooming celestial garden. The angel is a young feminine figure with very long flowing {argument name="hair color" default="golden blonde"} hair streaming across the scene, wearing a translucent white robe with delicate gold embroidery and a jeweled golden neckline, with two enormous feathered wings spread behind her. A brilliant sunlike halo and backlight glow directly above and behind the angel’s head, scattering gold sparks and blue-white light through the feathers. The angel gently places one hand on the nun’s bowed head in blessing while the nun clasps both hands in prayer. The nun wears a deep {argument name="nun habit color" default="navy blue"} veil and traditional habit with a white wimple, kneeling humbly and looking upward. Surround them with a dense dreamlike field of blue, violet, lavender, and peach flowers, drifting petals, glittering dust, and swirling strands of light. Composition is vertical portrait, close and intimate, with the angel occupying the upper left and center, the nun in the lower right, their hands forming the emotional focal point. Style: highly detailed painterly anime fantasy art, ethereal Catholic-inspired iconography, soft luminous brushwork, iridescent wings, dramatic rim lighting, warm gold highlights contrasted with cool blues and purples, romantic sacred atmosphere, no text, no watermark.

Create a realistic lifestyle fashion photograph of a young woman standing on a quiet sunlit city sidewalk in the morning, framed from mid-thigh to above the head in vertical portrait orientation. She has long, voluminous {argument name="hair color" default="soft pastel pink"} hair blowing slightly in the breeze and is lifting one hand to brush hair back near her right ear. Her face is intentionally obscured by a centered opaque square privacy block in a muted brown-pink skin-tone color, covering the entire facial area while leaving hair, neck, and body visible. She wears a fitted ribbed {argument name="top color" default="cream white"} sleeveless crop tank top and dark high-waisted denim jeans. Use natural candid street-photography styling with realistic body proportions, soft skin tones, and subtle fabric texture. The setting is an upscale urban street with pale building facades, shop windows on the right, sidewalk paving stones, and softly blurred storefronts and greenery receding into the background on the left. Lighting should feel like warm morning sunlight from the upper left, creating gentle highlights in the hair, soft shadows on the sidewalk, and a slightly airy overexposed film-photo look. Camera style: handheld eye-level portrait, shallow depth of field, 35mm documentary fashion photography, natural colors, no text, no logo, no watermark, no extra people, keep the pose casual and candid.

Create a photorealistic vertical smartphone selfie-style image of an adult woman cropped from just below the nose to mid-thigh, with no face or eyes visible. She is wearing a fitted {argument name="shirt color and style" default="white short-sleeve button-up blouse with a pointed collar"} and a {argument name="skirt style" default="high-waisted black pleated skirt"}. Her {argument name="hair color and length" default="straight dark brown shoulder-length hair"} falls over both sides of her chest, with a few strands resting on the blouse. The camera is held slightly above and close to the body, angled downward in a casual mirrorless selfie perspective, showing the left arm entering from the lower-left edge. Use soft indoor natural lighting, realistic fabric wrinkles, visible shirt buttons, subtle collar shadows, and shallow depth of field. Background is an indistinct indoor room with pale flooring and a muted curtain or wall, softly blurred. Keep the composition candid and realistic, vertical 9:16 aspect ratio, no text, no logo, no watermark, no exaggerated anatomy, no explicit nudity.

Create a cinematic dark-fantasy portrait of {argument name="two central warriors" default="a regal blonde woman in ornate silver full-plate armor and a towering shirtless barbarian man"} standing side by side in front of a grand medieval stone cathedral. The composition is vertical and symmetrical, waist-to-full-body framing, dramatic realistic lighting, high-detail photorealism, gritty fantasy-film atmosphere. Include exactly 2 main foreground figures: the left figure wears intricately engraved polished silver armor, gauntlets, greaves, a pale embroidered tabard, and a long cream-white cape; the right figure is extremely muscular, wet or oiled skin, long dark curly hair, black leather-and-fur war skirt, wrapped bracers, rugged boots, and a sheathed sword at his side. Place exactly 6 background guards in reflective steel helmets and armor, holding tall polearms or spears, arranged behind the main pair like a royal battle escort. Add deep navy banners with ornate gold heraldic patterns hanging on the cathedral walls, stone paving underfoot, and a ceremonial military formation feel. Cover the visible faces with plain opaque brown rectangular censor blocks: one on the left main warrior, one on the right main warrior, and one on the left background guard. Use {argument name="visual style" default="ultra-detailed photorealistic cinematic medieval fantasy"}, rich metallic reflections, high contrast, shallow depth of field, no modern objects, no text, no watermark.

Create a vertical 9:16 anime-style fashion illustration of a slender androgynous young adult model in Y2K streetwear, shown from upper thighs to head in a low-angle studio photo pose. The character has {argument name="hair color" default="messy golden blond"} medium-length layered hair with long face-framing strands, but the face is intentionally hidden by a centered opaque beige square censor block. Pose the model with the left hand tucked in a pants pocket and the right arm raised, fingers near the ear as if adjusting an earring, giving a cool runway attitude. Include exactly 7 visible fashion elements: 1 oversized pale blue denim jacket worn open and slipping off one shoulder, 2 cropped plain white T-shirt exposing the midriff, 3 teal-blue low-rise ripped jeans, 4 black waistband with tiny abstract white markings but no readable brand logo, 5 chunky silver ear jewelry, 6 black bracelet or wrist accessory, and 7 small gold chain detail hanging from the jeans pocket. Make the jeans have exactly 3 major torn knee/thigh openings: one large rip on the left thigh, one large rip on the right thigh, and one smaller rip lower on the right leg. Use crisp high-detail Japanese anime rendering with glossy fabric highlights, sharp ink lines, warm skin shading, dramatic sunlight from above, and a bright white photography-studio background with visible black light stands, ceiling rigging, and round studio lights. Add a faint white social-media watermark across the upper hair area reading {argument name="watermark text" default="Yume@yumeK"}. Avoid any recognizable brand logos or real fashion trademarks.

Goal: Create a luxurious photographed paper-craft pop-up business card for an original anime character, like an official dimensional calling card with a full storybook world unfolding from the card.

Canvas: Vertical 4:5 image, realistic close-up product photography on a soft beige tabletop, shallow depth of field, warm natural light from the left, blurred antique mirror/gold ornament in the back left and cream roses in the back right.

Main subject: A freestanding layered pop-up card in ivory, dusty blue, pale silver, and pearl white. The card forms an arched miniature room with ornate laser-cut paper layers, filigree borders, floral decorations, and a seated anime girl in the center. The character is {argument name="character name" default="Nozomu"}, a delicate blonde girl with long straight hair, soft gray-blue eyes, faint blush, and a melancholy gentle expression. She wears a muted blue dress or oversized blouse with white frilled trim, layered necklaces, earrings, and a blue headband with white flower hair ornaments and ribbons. She sits curled up with one knee raised, one hand resting near her cheek, facing slightly right but looking toward the viewer.

Pop-up construction and countable elements: Use exactly 5 major vertical paper layers: 1 front nameplate base, 1 floral sofa/ledge layer, 1 character cutout layer, 1 ornate arched window layer, and 1 rear blue-and-ivory room backdrop. Add exactly 1 cream chandelier centered at the top. Add exactly 8 hanging crystal/star ornaments around the arch: 2 star charms, 3 diamond-shaped crystals, and 3 small teardrop crystals. Add exactly 2 decorative side lattice panels, one on each side. Add exactly 4 visible flower clusters: front left blue-white cluster, front right blue-white cluster, rear left pale cluster, rear right pale cluster. Add exactly 2 small pillows, one on each side behind the character.

Text content: On the large front plaque, write the name in elegant serif capitals: {argument name="main name text" default="NOZOMU"}. Under it, write {argument name="subtitle text" default="Silent Blue Muse"}. Under that, in smaller type, write {argument name="small descriptor text" default="official pop-up business card"}. On a tiny bottom label, write {argument name="serial number" default="No. 0001"}. On a small vertical side tag at left, include the Japanese hiragana text 「のぞむ」 in black ink.

Visual style: Highly detailed anime illustration printed on layered die-cut paper, elegant rococo and Victorian stationery motifs, pale blue floral fantasy aesthetic, thin silver linework, pearl accents, soft shadows between paper layers, realistic paper thickness, glossy highlights on crystal ornaments, refined handmade craft quality.

Constraints: The image should look like a real physical pop-up card being photographed, not a flat poster. Keep all text legible and centered on the plaques. Do not add extra characters, extra labels, watermarks, logos, hands, or packaging.

Goal: Using REFERENCE_0 as the base photo, transform it into a warm Japanese hand-drawn watercolor-and-colored-pencil scrapbook journal page while preserving the same subject, pose, school-uniform jacket, seated convenience-store window composition, rainy night mood, and anonymized/obscured face.

Style transformation: Convert the realistic photo into a cozy illustrated “late night journal” spread on beige grid notebook paper, with soft ink outlines, muted watercolor textures, visible paper grain, rounded page corners, tape scraps, doodles, arrows, hearts, raindrops, and handwritten Japanese notes. Make it feel nostalgic, cute, and slightly messy like a personal travel diary.

Scene changes and additions: Replace the plain convenience-store snack setup with a richer Japanese convenience-store meal scene. Add exactly 7 visible food/drink items on and near the counter: 1 oden cup with large Japanese label, 1 melon bread held in the subject’s hand, 1 bottled milk tea, 1 wrapped onigiri, 1 packaged snack bag, 1 small plate with an extra melon bread, and 1 small receipt/slip. Outside the window, emphasize a rainy Shinjuku night skyline with blurred neon and Tokyo Tower-like red landmark lights.

Journal layout elements: Add exactly 10 scrapbook/annotation elements around the illustration: 1 large handwritten Japanese headline at top left, 1 small English phrase “late night journal,” 1 taped label reading “Daily Konbini Life,” 1 top-right note card, 1 left-side handwritten thought box, 1 left checklist box for today’s purchases, 1 small convenience-store tote-bag doodle, 1 bottom-left Polaroid photo labeled “Shinjuku Rainy Night,” 1 bottom-right memo checklist card, and 1 speech bubble near the subject. Use arrows and tiny doodles to connect notes to the food and window view.

Text: Use mostly handwritten Japanese with a few casual English accents. Main headline text: {argument name="headline text" default="コンビニ日和。"}. Top-left English phrase: {argument name="journal phrase" default="late night journal"}. Taped label text: {argument name="tape label" default="Daily Konbini Life"}. Polaroid caption: {argument name="polaroid caption" default="Shinjuku Rainy Night"}. Memo card title: {argument name="memo title" default="MEMO"}.

Constraints: Keep the composition based on the reference rather than inventing a new character angle. Preserve the anonymous blurred/blank face. Do not make it photorealistic; it must look like an illustrated notebook page. Avoid extra characters, logos, watermarks, or clean digital UI styling.