GPT Image 2 prompts

anime character reference sheet,
full body character design sheet,
official visual development art,
professional game concept art,
luxury anime RPG UI,
anime model sheet,
turnaround sheet,
expression sheet,
weapon sheet,
character profile board,

Create a {argument name="style" default="Pixar 3D"} style character design sheet. Clean white background. Two characters side by side with a clean dividing line. Bold brushstroke-style title at the top: {argument name="title" default="BARRY vs THE POPCORN"}. Subtitle beneath: {argument name="subtitle" default="One bucket. Zero timing."}

Create an English infographic catalog poster showing {argument name="era" default="1990s"} anime character styles in a clean labeled grid layout, each panel featuring a different archetype and visual language. Use 16:9 aspect ratio.

Goal: Create an absurd maximalist vertical advertising poster for {argument name="brand phrase" default="TACO SUPREME EXTREME"}, rendered like a hyper-saturated 1990s extreme-sports food commercial mixed with sci-fi fantasy.

Canvas: Tall vertical 9:16 poster, ultra-detailed, glossy, high contrast, neon magenta and orange sky, cinematic depth, sharp foreground and densely packed background.

Main subject: One gigantic anthropomorphic hard-shell taco dominates the center, flying diagonally upward on one flaming skateboard. The taco shell is golden and crunchy, stuffed with seasoned beef, lettuce, tomatoes, cheese, and peppers. It wears one pair of oversized black sunglasses and shoots two intense rainbow laser beams from the lenses toward the right edge. Around its neck is one thick gold chain with one pendant reading “TACO SUPREME.” Add small legs or shoes under the taco to make it feel like a living mascot.

Top title: In the upper left, place three stacked words of giant metallic chrome 3D text: “TACO” on top, “SUPREME” below, and “EXTREME” in flaming italic script beneath. Surround the title with orange fire, sparks, hot highlights, and black drop shadows so it looks like a loud energy-drink logo.

Background and surreal elements: Include exactly 6 major fantasy set pieces: 1 erupting volcano in the lower left background, 1 sneaker-shaped UFO in the upper right, 1 sad pink brain-like cloud under the UFO crying streams of blue liquid, 1 sprawling futuristic city made of stacked burgers and fries, 1 giant tilted soda bottle pouring dark fizzy soda at the lower right, and 1 glowing magenta river or street running through the city. The sneaker UFO should be colorful, glossy, covered in cyan and pink lights, and projecting a blue tractor beam onto the crying cloud.

Crowd and city details: Fill the lower half with a miniature burger metropolis. Use many burger towers, fry skyscrapers, neon signs, and tiny suited yellow mascots wearing black sunglasses. Make the mascots look like secret agents standing on burgers, boats, rooftops, and streets; keep them secondary and tiny compared with the main taco. Add at least one small rectangular sign in the city that repeats the brand phrase.

Atmosphere: Add abundant floating diamonds, colorful confetti rectangles, sparks, glitter, embers, and food particles throughout the sky. Use explosive pink powder clouds near the volcano and a glowing sunset horizon behind the city.

Visual style: Hyperreal 3D illustration, glossy toy-like surfaces, chrome typography, dramatic rim lighting, lens flares, saturated rainbow beams, fire trails, chaotic composition, extremely detailed, playful, surreal, ridiculous, premium poster finish.

Constraints: No photoreal humans, no plain background, no minimalism, no muted colors. Keep all visible text limited to the brand phrase and its repeats. Make the composition feel loud, funny, and impossibly energetic.

Create a minimalist luxury fragrance product shot on a clean white studio background. Center a single futuristic perfume bottle shaped like a perfectly round polished chrome sphere with a flat circular base and a short cylindrical chrome spray cap on top. The bottle should look highly reflective and mirror-like, with smooth black-and-white studio reflections curving across the spherical surface, a dark soft central reflection, crisp metallic rim highlights, and subtle grounding shadow beneath. The sprayer has a tiny nozzle aperture visible near the top front, with a vertical black reflective strip running down the cap. Add engraved metallic text across the front of the sphere: {argument name="fragrance name" default="ATOMIUM AIR™"} in thin uppercase modern sans-serif lettering, and a smaller tagline beneath it reading {argument name="tagline" default="Brussels, bottled in the future"}. The composition is symmetrical, premium, quiet, and hyper-realistic, like a high-end cosmetics advertisement inspired by Belgian futurism, Art Deco chrome, rain-slick Brussels, and spacecraft-like design. Use only one bottle, one cap, one base, and two visible text lines; no people, no props, no colored background, no extra labels, no watermark.

Using REFERENCE_0 as the product and lifestyle-photo base, transform the casual shoe photo into a polished Nike-style product advertising campaign for {argument name="product name" default="Nike Mind 001"}. Keep the same white slip-on recovery shoe concept, the orange massage nodes in the footbed, and include the original deck photo as one collage element, but redesign everything as a clean editorial brand layout.

Create exactly 3 stacked horizontal poster panels on one vertical canvas, separated by thin borders:

Panel 1: Hero launch ad. Show a large floating cutout render of one shoe angled upward, with a beige desert/rock landscape strip on the right and minimal paper-texture collage pieces. Add a large bold headline reading “Nike” and “mind 001”. Add the subheadline {argument name="recovery tagline" default="Recovery for Body. Calm for Mind."}. Include small product-system text in the upper left, a small handwritten oval note reading “mental reset.”, three simple circular feature icons labeled “LIGHTWEIGHT COMFORT”, “MASSAGE FOOTBED”, and “EASY SLIP-ON”, plus a small product label card and the footer phrase “MOVE LESS. FEEL MORE.”

Panel 2: Lifestyle collage ad. Use the reference photo as the main rectangular photo in the upper center/right, preserving the feet-on-deck viewpoint. Surround it with cutout product detail photos: one close-up of the perforated upper, one angled view of the orange massage footbed, and one close-up of the wavy cushioned sole. Add a bold headline on the left reading {argument name="middle headline" default="Made to Feel Better."}. Add a small handwritten note at the top left reading “Every step, a soft reset.”, a paper note at the top right reading “Take it easy. You’re doing great.”, a small tab reading “DETAILS 001”, and handwritten annotation arrows for “Breathable upper Soft support. All day calm.” and “Cushioning that moves with you.”

Panel 3: Multi-angle product poster. Show exactly 3 floating cutout views of the shoe on a white/off-white background: one angled top view near the top center, one side/top view near the lower left revealing the orange footbed, and one sole/side view on the right. Add large stacked typography on the left reading “Mind Set. Go Light.” with “Light.” in orange. Include small Nike Mind Series text, a circular handwritten note reading {argument name="handwritten note" default="Simple. Soft. Essential."}, a vertical barcode label on the right reading “MIND 001” and “YOUR DAILY RESET.”, and an orange call-to-action button reading {argument name="button text" default="SHOP NIKE MIND 001 →"}.

Style: premium minimalist sneaker campaign, off-white paper background, soft shadows, cutout collage aesthetic, black/gray typography with orange accent highlights, subtle taped-paper textures, handwritten red/orange pencil annotations, clean modern sans-serif type. Make the shoes look like high-end studio product renders while still clearly derived from the reference shoe. Avoid adding extra panels, extra shoes beyond the specified product views, watermarks, or unrelated branding beyond Nike-style marks and text.

Goal: Create a hyper-realistic vertical cinematic soda advertisement poster for {argument name="beverage brand" default="Dr Pepper"}, styled like premium commercial beverage photography with dramatic red lighting and a retro-night atmosphere.

Canvas: Vertical 2:3 poster, high-resolution, glossy commercial finish, dark black-and-crimson color palette, shallow depth of field, wet reflective street, neon glow, strong contrast, realistic skin and fabric texture.

Main subject: A fashionable young woman stands in the left foreground in a confident commercial pose, leaning near a shiny red classic car while extending one hand toward the camera holding a cold red soda can. She has messy tied-back dark hair, natural realistic skin, soft makeup, and elegant earrings. Her face is intentionally covered by a centered plain rectangular skin-tone censor block, while the rest of her body remains detailed. Outfit pieces: exactly 5 visible outfit elements — burgundy fitted blazer, black crop top, black high-waisted jeans, black heeled boots, and earrings.

Product hero: The soda can is large and close to the lens in the right-center foreground, covered in condensation droplets, metallic red, with a clear white {argument name="can logo text" default="Dr Pepper"} logo. The woman’s fingers wrap around the can realistically. From the open top of the can, a dramatic swirling wave of dark red soda explodes upward and arcs across the top of the poster like a liquid ribbon. Include exactly 5 visible ice cubes suspended inside the splash, plus many small droplets and bubbles.

Background and scene: Set the poster outside a glowing retro American diner at night. The diner is warm-lit with glass windows, silhouettes of patrons, and a rooftop neon sign reading {argument name="diner sign text" default="DINER"}. The soda splash visually contains a miniature cinematic reflection/scene of the diner, city skyline, and classic cars inside its liquid curve. Include exactly 3 visible cars: one red classic car in the lower-left foreground, and two classic cars parked near the diner in the upper background. Add a large blurred neon diner sign on the far left and a roadside billboard on the right.

Visible text content: Include exactly 6 main readable text elements: 1) the can logo "Dr Pepper", 2) small can text "Est. 1885", 3) circular can seal "23", 4) the diner rooftop sign "DINER", 5) the right-side billboard logo "Dr Pepper" with "23" below it, and 6) the bottom headline {argument name="headline text" default="BOLD FLAVOR"} with the tagline {argument name="tagline text" default="NOT ORDINARY. NEVER BORING."}. Make the headline huge, bold, distressed white block lettering in the lower-right corner; make the words "NEVER BORING." red.

Visual style: Hyper-realistic cinematic beverage campaign, luxury soda commercial poster, dramatic specular highlights, glowing red neon, rain-slick asphalt, volumetric night lighting, sharp product focus, realistic condensation, high-impact advertising composition.

Constraints: Keep the product can as the clearest focal point, preserve the face-obscuring rectangle, avoid extra brands beyond the chosen beverage, avoid misspelled text, no watermark, no extra poster borders.

Create a photorealistic product mockup of a cute Japanese tear-off daily desk calendar standing on a warm wooden table in soft morning light. The calendar is an A-frame tabletop block with thick cream paper pages and a textured off-white binding strip along the top, viewed from a slightly front-left angle. The visible page has a pastel watercolor anime design in mint, cream, peach, and pale blue, with delicate paper texture, scalloped frames, tiny daisies, and a small bow ornament near the top. On the right half of the page, illustrate one sweet original anime girl: blonde peach-pink curly twin-tail hair, large sparkling aqua eyes, rosy cheeks, a gentle smile, chin resting on both hands, wearing a mint green dress with tiny white and yellow flowers, a frilly blue-white collar, scrunchie bracelets, and blue ribbon bows plus small flower clips in her hair. On the left half of the page, include exactly five text elements: top small label 「うちの子日めくり」, large date “{argument name="date" default="5/24"}”, weekday “{argument name="weekday" default="日曜日"}”, character name “{argument name="character name" default="のぞむ"}”, and a rounded cloud message box reading “{argument name="daily message" default="あなたのペースでね"}”. Use elegant Japanese typography, with the date in large teal serif numerals and the other text in soft brown and teal. Surround the calendar with cozy desk props: exactly one mint floral pen lying diagonally in the foreground, one glass vase with white and blue flowers on the left, one white lace cloth under the pen, one softly blurred glass container on the right, and small out-of-focus white blossoms in the lower right foreground. Keep the mood gentle, handmade, feminine, airy, and nostalgic, with shallow depth of field, warm highlights, realistic shadows, and no watermark or extra text.

Create a cinematic widescreen anime night scene in a dense old East Asian city street after rain, with a lone young man {argument name="main character" default="a blond young man in a sleeveless black outfit"} running away from the viewer toward an ornate temple gate at the end of the street. Use a low, centered camera angle from behind the runner, making him the dominant silhouette in the foreground with one leg lifted mid-stride and arms pumping. The setting is a narrow wet alley lined with weathered shopfronts, shuttered storefronts, vertical signboards with Chinese-style characters, hanging wires, potted plants, and glowing neon. At the far end, place an elaborate traditional temple facade and gateway with layered curved roofs, carved details, and glowing signage. The ground is slick with rain and reflects vivid teal, cyan, magenta, and hot pink lights. Include exactly 8 visible human figures: 1 central running protagonist in the foreground, 2 startled men on the left side of the street, 1 older bald man standing near the left storefront, 2 men near the temple entrance, 1 man on the right side of the entrance, and 1 small distant silhouette under the archway. Include exactly 5 prominent round lanterns on the right side, glowing pink and orange, plus smaller scattered lantern glows in the distance. Mood: tense, breathless, mysterious, slightly sacred and dangerous, like a chase through a neon temple district at midnight. Visual style: highly detailed 2D anime background painting, moody cyberpunk color grading, deep shadows, painterly linework, cinematic composition, atmospheric haze, high contrast rim lighting, reflective pavement, no photorealism. Color palette: {argument name="color palette" default="deep teal shadows, cyan highlights, neon magenta, hot pink lantern light, dark indigo sky"}. Aspect ratio: {argument name="aspect ratio" default="wide 2.35:1 cinematic frame"}. Avoid readable modern branding, avoid cars, avoid daylight, avoid extra foreground characters, and keep the protagonist's face hidden from view.

Create a cinematic wide 2.39:1 anime-style night scene inside a moody neon bar or rainy street-facing lounge, using deep indigo shadows and electric cyan-blue highlights with a few violet accents. In the exact center foreground, show {argument name="main character" default="a young adult man"} from the chest up, facing the viewer, wearing a dark open jacket over a pale shirt; his short messy {argument name="hair color" default="black"} hair is rim-lit in blue, his shoulders fill the lower half of the frame, and his face is deliberately obscured by a large nearly black square censor block with a subtle purple gradient, covering the entire facial area while leaving hair, ears, neck, and shoulders visible. The atmosphere should feel tense, quiet, and slightly startling, like a noir anime key visual. Behind him, include exactly 9 background people, for 10 total human figures: 3 dim silhouettes seated or leaning at a bar booth on the far left, 1 tall standing silhouette just left of center behind the main character, and 5 soft silhouettes on the right side near glass windows, including two people conversing, one holding a drink with a straw, one standing farther back, and one small figure at the far right. The background is a blurred neon interior with glowing vertical blue light strips, glass panels, vague signage, and wet reflective flooring with puddle-like reflections in cyan. Use high-contrast cel-shaded anime lighting, painterly digital background detail, shallow depth of field, strong rim light on the central figure, and no readable text, logos, or watermark. The overall palette should be {argument name="color palette" default="midnight blue, cyan neon, deep purple, black"}, and the scene mood should be {argument name="mood" default="silent, uneasy, holy-night neon melancholy"}.

Create a cinematic anime-style night scene in a neon-lit urban shopping district, with {argument name="character name" default="an unnamed young man"} as the single main subject in the foreground. Use a wide horizontal 2.35:1 composition, viewed from street level after rain, with glossy wet pavement reflecting cyan, teal, violet, and magenta lights. The main subject is a slim young man in his late teens or early twenties, shown from chest up and slightly right of center, wearing a pale button-up shirt with a small chest pocket; his head is turned slightly left, his short dark hair forms a soft silhouette, and his entire face is deliberately obscured in deep shadow with no visible facial features, creating a mysterious, breathless atmosphere. Behind him, show exactly 9 secondary human figures: 4 larger dark silhouettes in the midground, including one person carrying a bag, and 5 tiny distant pedestrians scattered along the crossing and sidewalks. The background is a dense modern city intersection with tall glass buildings, stacked illuminated windows, storefront signs, vertical light panels, and a traffic signal glowing pink-red. Make the architecture feel like a rainy Tokyo nightlife district without readable signage. Lighting should be moody and high contrast, dominated by {argument name="neon palette" default="cyan, teal, deep blue, violet, and hot magenta"}; the foreground character should be mostly navy and purple shadow with thin turquoise rim light on the shirt collar and shoulder. Use expressive painted anime background art, subtle grain, soft bloom, reflective puddles, and a quiet melancholic mood inspired by {argument name="mood reference" default="a silent holy-night urban mood board"}. Avoid text, logos, watermarks, clear facial details, daylight, or cheerful colors.

Create a cinematic anime-style educational landscape illustration of a glacial fjord, designed as a widescreen 16:9 slide. Show a dramatic split-level cross-section where the upper half is above the waterline and the lower half is underwater, with the water surface forming a crisp horizontal line across the middle. Above water: towering steep mountains on both sides of the fjord, snow-capped peaks in the distance, a bright blue sky with scattered white clouds, small floating icebergs, and a glacier visible at the far end of the inlet. On the right side, emphasize a massive curved cliff face with clear layered sedimentary rock strata bending in broad arcs; on the left, darker rugged mountains with patches of green vegetation and snow. Below water: transparent turquoise-blue fjord water revealing submerged valley walls, rounded stones, rocky slopes, glacially carved terrain, and continued layered rock formations, with soft underwater light rays and depth haze. Add a subtle dark translucent gradient panel across the bottom third for readability. Place exactly one centered Japanese educational text block in large white sans-serif characters with a slight shadow, broken into four lines, using this text: {argument name="caption text" default="この地域は、美しいだけではなく、地球のしくみを学ぶ場所でもあります。海底のモレーンや、氷河に磨かれた岩肌、地殻の動きの痕跡などが残されています。フィヨルドは、過去の氷河活動と今も続く地形変化を見せてくれる、野外の地球科学博物館なのです。"} Visual style: high-quality Japanese anime background art, painterly but detailed, naturalistic lighting, crisp geology, serene atmosphere, vibrant blues and greens, no people, no boats, no logos, no watermark.

Goal: Create an antique parchment anime character analysis chart for {argument name="character name" default="Naruto"}, styled like an ancient ninja manuscript and role anatomy poster.

Canvas: Vertical 4:5 poster on aged tan paper with burnt, torn, darkened edges, ink stains, cracks, hand-drawn borders, weathered texture, and symmetrical decorative scrollwork. Use a sepia, black ink, orange flame, and blue chakra palette. The whole page should look like a detailed archival character dossier from a fantasy shinobi world.

Top header: Large bold black brush-calligraphy title centered, reading {argument name="main title" default="【鸣人】"}, with a smaller subtitle plaque underneath reading character anatomy chart in Chinese-style lettering. Add two large decorative spiral emblems near the top left and top right, surrounded by smoky cloud patterns, plus small red seal stamps. Put narrow vertical side plaques near the upper corners with Chinese-style labels.

Main layout: Divide the poster into five tall vertical manuscript columns plus a right-side detail strip, each framed with thin black hand-inked borders and bracketed Chinese-style section headers. The five main columns are: 1) expression comparison, 2) appearance outfit, 3) ability structure, 4) desire core, 5) detail magnification and evolution path.

Column 1, expression comparison: Show exactly 2 stacked character portrait panels. The upper panel is a calm front-facing bust of the anime ninja in a black-and-orange jacket with spiky blond hair and whisker cheek marks; the face area may be softly blurred or simplified. The lower panel is an intense transformed fiery fox-chakra bust, with glowing orange aura, fierce eyes, whisker marks, and flame-like energy rising around the head. Add small bracket labels beside both portraits.

Column 1 lower section, trait icons: Show exactly 6 circular trait icons arranged in a 2 by 3 grid, each in its own small bordered square with a tiny label. The six icons are: clone silhouettes, flame drop, spiral vortex, seated sage silhouette, nine-tailed fox flame, and protective core emblem.

Column 2, appearance outfit: Full-body front view of the character standing upright, occupying most of the column. He has {argument name="hair color" default="spiky blond hair"}, a black forehead headband, black-and-orange zip jacket, orange cropped shinobi pants, dark sandals, black ankle wraps, bandage wrap on one thigh, and a small pouch on one leg. Use clean anime line art but weathered ink texture. Add exactly 8 small annotation labels with leader lines pointing to outfit pieces: headband, hair, collar/zipper, jacket body, sleeve patch, pants, pouch, sandals.

Column 3, ability structure: Draw a glowing anatomical chakra diagram of the same standing body as a transparent golden outline. Show visible internal energy channels, circular chakra nodes, and a bright blue spiral core at the chest. Use exactly 7 labeled callouts pointing to energy features: crown chakra, clone technique, spiral orb, sage path, tailed beast force, sealing core, and nine-tailed power. Surround the body with smoky blue and gold energy lines.

Column 4, desire core: Create a tall dramatic illustration of an orange nine-tailed fox spirit made of fire and swirling chakra, looming from top to center. Below it, show a small dark silhouette of the hero facing the energy storm. Include exactly 5 labeled callouts: fire dream, spiral heart, unwillingness to yield, protective core, and inner beast. The column should feel the most chaotic, with dense flame curls and a large vortex near the lower center.

Right detail strip: Include two stacked subsections. The upper subsection is detail magnification with exactly 4 horizontal close-up boxes: metal forehead protector with spiral symbol, one blue eye close-up, hands holding a blue spiral energy orb, and golden chakra-flow pattern. The lower subsection is evolution path with exactly 5 stacked small portrait thumbnails connected by downward arrows: academy child, lower-rank ninja,疾风-era teen, sage mode, and nine-tail mode. Keep the thumbnails small and manuscript-like, with face areas simplified if necessary.

Bottom border decorations: Add exactly 5 bottom decorative elements: a large blue-black spiral vortex at bottom left, a rolled cylindrical scroll near the lower left-center, a central spiral medallion plaque, a kunai dagger angled beside a sealing paper tag, and a blue wave ornament at bottom right. Add a small centered bottom plaque with Chinese-style text, plus vertical plaques in the lower left and lower right corners.

Visual style: Highly detailed anime illustration mixed with ancient Chinese/Japanese scroll aesthetics, black ink outlines, brush calligraphy, faded red seals, hand-numbered chart feeling, ornate but readable compartments, grunge paper texture, subtle watercolor wash, glowing orange and blue magical energy. Use no modern UI elements, no photorealism, no clean white background, and no extra characters beyond the single character and symbolic fox spirit.

Create a vertical 2:3 high-fashion editorial photo of an adult female model wearing a dramatic liquid dress made from glossy pink paint. The model stands centered in a studio against a plain light gray wall, captured from mid-thigh upward with her body slightly turned and one arm extended to the right. Her face is intentionally hidden by a flat opaque square censor block in a skin-tone beige color, while her long wavy hair is {argument name="hair color" default="blonde at the crown fading into vivid hot pink lengths"}. The dress should look like a living splash of {argument name="liquid dress color" default="bubblegum pink and translucent pale pink"}: an off-shoulder bodice with a strapless neckline and a huge swirling skirt formed from sheets of liquid, ribbons of paint, droplets, beads, splashes, and thin tendrils frozen in midair. Count and emphasize 5 main sweeping liquid shapes: one wide horizontal band across the chest, one large curling splash rising on the left, one broad skirt sheet flaring to the lower left, one bright magenta skirt panel flaring to the right, and one lower back wave with dripping edges. Add many small suspended droplets around the figure, especially along the left splash and right side. The model wears sparkly jewelry including a choker necklace, bracelets, and small hair ornaments. Use direct on-camera flash, glossy wet highlights, slight film grain, hard shadows on the wall, a glamorous 2000s music-video mood, realistic skin and fabric-liquid physics, high detail, dynamic frozen motion, no visible text, no watermark.

Goal: Create a side-by-side anime illustration comparison demonstrating subsurface-scattering translucent light, with the same character shown in two vertical panels: a normal version on the left and an enhanced glowing version on the right.

Canvas: Wide 16:9 social-media comparison image, about 1200×675 px, split exactly down the center by a thin vertical white divider line. Use a dark blue nighttime color palette with strong contrast between the left “Before” and right “After” lighting.

Layout: Two equal panels. Top left label reads “Before” in bold white text with a black shadow. Top center headline reads “プロンプト比較：SSS透過光” in large bold white Japanese text with a dark outline. Top right label reads “After” in bold white text with a black shadow. Bottom left caption reads “通常生成” and bottom right caption reads “SSS透過光”, both in bold white Japanese text with black shadow. Keep exactly 5 text elements: 1 Before label, 1 Japanese headline, 1 After label, 1 bottom-left Japanese caption, and 1 bottom-right Japanese caption.

Subject details: Show the same youthful anime boy in both panels, upper body portrait, three-quarter pose facing slightly left, slim build, pale skin, short fluffy lavender-silver hair with many soft strands, visible ear, delicate neck and hand, wearing a loose white robe or pajama-like shirt with a dark navy star-patterned collar. Add celestial accessories on the chest: exactly 3 visible hanging ornaments, including a crescent moon brooch, a small star charm, and a dangling chain detail. His right hand lightly holds the collar near the chest. The face is intentionally covered by a large opaque square privacy block centered over the head in both panels; use a muted lavender-gray square on the left and a warmer pink-lavender square on the right.

Background: Both panels show the same dreamy bedroom/window scene at night: tall window frames, dark navy starry sky, a crescent moon visible behind the character, sheer curtains, tiny sparkles and soft bokeh. The left panel should feel cooler, flatter, and less luminous. The right panel should feel more cinematic and premium, with warm rim light and glowing particles.

Lighting comparison: Left panel is normal generation with soft cool moonlight, subdued shadows, minimal translucency, hair edges mostly matte, ear and cheek not glowing. Right panel adds dramatic SSS translucent light: warm pink-orange light appears to pass through the ear, cheek area, fingertips, and hair edges from behind, creating a halo rim around the hair, warm bloom on the skin, luminous outline on the shoulder and hand, and more depth. Emphasize transparency, dimensionality, and high-quality anime rendering on the right without changing the character pose.

Visual style: High-end Japanese anime illustration, soft painterly rendering, clean line art, glossy hair highlights, dreamy blue-purple night atmosphere, subtle magical sparkle effects, crisp readable typography, polished comparison graphic suitable for an AI art tutorial post.

Customizable variables: Use {argument name="headline text" default="プロンプト比較：SSS透過光"}, {argument name="character description" default="youthful anime boy with fluffy lavender-silver hair"}, {argument name="left caption" default="通常生成"}, {argument name="right caption" default="SSS透過光"}, and {argument name="lighting effect" default="warm subsurface-scattering translucent light through cheeks, ears, and hair edges"}.

Constraints: Keep the two character portraits nearly identical in pose and outfit. Include exactly 2 large face-covering squares, one in each panel. Do not add extra panels, extra captions, watermarks, logos, or unrelated objects.

Goal: Create a square comparison sheet showing two different storybook-style image generations from the same rainy dragon prompt, with the top half labeled as a warm red-raincoat version and the bottom half as a hydrangea-blue rainy version.

Canvas: Square 1:1 white background, like a clean social media comparison post. Divide the canvas into two horizontal halves separated by a centered Japanese caption line. Use hand-drawn panel borders with slightly uneven ink lines and softly rounded corners.

Top section: Label the upper-right corner in small red handwritten text: “Nano Banana2”. Arrange exactly 3 panels: 1 wide rectangular panel at upper left showing a child in a red raincoat struggling with a beige umbrella in heavy wind and rain, with dark storm clouds and bent trees; 1 wide rectangular panel below it showing a side-profile child in the same red raincoat holding the umbrella against a mostly blank white background; 1 large tall rounded rectangular panel on the right showing a friendly orange dragon bending down under the rain toward the child in a red hooded raincoat, the dragon’s wings forming a warm shelter-like shape. The child’s face may be softly blurred or simplified. Use warm orange, red, and ochre as the dominant colors.

Center caption: Add one centered line of small Japanese text between the halves: “ともに温かみのある、絵本のようなイラストという指定”.

Bottom section: Label the lower-left corner in small pastel purple handwritten text: “GPT Image2”. Arrange exactly 3 panels: 1 small panel at bottom-left top showing a little girl in a yellow raincoat standing in heavy rain under a blue umbrella, surrounded by wet pavement, streetlamp glow, and purple-blue hydrangeas; 1 small panel directly beneath it showing the same girl looking upward with a small white question mark near her head; 1 large wide panel on the right showing the girl smiling up at a gentle white-blue dragon in the rain, with the dragon lowering its head toward her and spreading a translucent wing, hydrangeas blooming on both sides, reflections on the wet ground, and a magical blue twilight atmosphere.

Subject details: The main child is {argument name="main character" default="a young child or little girl"}; the top outfit is {argument name="top outfit color" default="red hooded raincoat"}; the bottom outfit is {argument name="bottom outfit color" default="yellow raincoat with blue skirt"}; the dragon theme is {argument name="dragon colors" default="orange dragon above and white-blue dragon below"}; the weather theme is {argument name="weather mood" default="heavy rain during the rainy season with hydrangeas"}.

Visual style: Warm children’s picture book illustration, watercolor and colored pencil texture, soft ink outlines, visible brush granulation, gentle expressions, cinematic rain streaks, cozy emotional storytelling. Top half feels windy, warm, and orange-red; bottom half feels moist, luminous, blue-purple, and hydrangea-themed.

Constraints: Use exactly 6 illustrated panels total, exactly 2 model labels, and exactly 1 centered Japanese caption line. Do not add extra UI, logos, watermarks, or additional text. Keep the composition readable as a side-by-side model comparison sheet.

Goal: Create a monochrome manga magazine cover / chapter splash page for {argument name="series title" default="なるようになるまで"}, with the episode label {argument name="episode label" default="第4話"}. The page should feel like a quiet, dialogue-free slice-of-life urban story, highly detailed and atmospheric.

Canvas: Vertical manga page, approximately 2:3 aspect ratio, black-and-white ink drawing with grayscale screentone, dense crosshatching, rain reflections, high detail, no color.

Layout: A collage composition with exactly 7 visible story scenes. The largest central scene shows the main character standing full-body in a narrow rainy Tokyo alley, surrounded by tall buildings, utility poles, shop signs, wet pavement, a manhole cover, and a distant broadcast tower. Around it, arrange 6 smaller rectangular vignette panels: 1 upper-left cramped bedroom scene, 1 upper-right convenience-store cashier scene, 1 middle-left tabletop budgeting scene, 1 middle-right laundromat waiting scene, 1 lower-left rooftop city-view scene with pigeons, and 1 lower-right nighttime bedroom scene.

Subject details: The same teenage manga girl appears in all 7 scenes. She has short dark bobbed hair with a small ahoge curl, large over-ear headphones, a loose oversized hoodie jacket with sporty sleeve tape, a pleated skirt in the central full-body scene, chunky futuristic sneakers, and a tired introverted posture. Her face should be intentionally obscured with a soft rectangular gray blur in the largest central figure and the closer vignette figures, creating an anonymous censored look while preserving the hair and headphones silhouette. Use {argument name="jacket marking" default="893"} as a small printed number or patch on the jacket in several scenes.

Exact scene count and contents: Scene 1, central full-body alley portrait: the girl stands hands in pockets, wet street reflections, dense Japanese urban signage, overhead wires, and a distant tower. Scene 2, upper-left bedroom: she sits on a messy bed looking at a phone, surrounded by clothes, bottles, posters, small table clutter, and a window. Scene 3, upper-right convenience store: she works at a register with shelves of canned goods and products behind her, wearing the same headphones and jacket. Scene 4, middle-left tabletop: she sits at a table with a cup-noodle container, bottled tea, coins, calculator, receipts, a wallet, and papers, suggesting budgeting or counting change. Scene 5, middle-right laundromat: she sits alone beside round washing machines, head down, with a large bag nearby. Scene 6, lower-left rooftop: she sits with a cup noodle and chopsticks overlooking a dense city skyline, with exactly 3 pigeons visible nearby and a drink bottle on the ground. Scene 7, lower-right bedroom at night: she lies awake in bed under blankets, city lights visible through the window, a digital clock on the wall or shelf, cup, phone, remote, small devices, and clutter on a bedside table.

Typography: Put the large brush-calligraphy title across the very top, bold black ink strokes with splatter and rough dry-brush texture. Place the smaller episode label below the left side of the title. Keep the typography integrated like a manga chapter opener.

Visual style: Detailed seinen manga line art, gritty urban realism, rainy Tokyo backstreets, cinematic composition, layered panels, ink splatter, screentone shading, high contrast, melancholy mood, no speech bubbles, no dialogue, no color, no watermark. Use {argument name="overall mood" default="quiet melancholy after rain"} as the emotional tone.

Using the provided reference image as the subject, transform the madeleine into a clean technical blueprint drawing rather than a photo. Create a vertical engineering-sheet layout on a dark blue grid background with thin white linework, a double border, measurement arrows, labels, and schematic outlines that preserve the madeleine’s shell-like oval form.

Goal: Produce a blueprint titled {argument name="title text" default="MADELEINE"} with the subtitle {argument name="subtitle text" default="TECHNICAL BLUEPRINT"}.

Layout: Include exactly 4 technical drawing panels derived from the reference pastry: 1 large top view at the upper right, 1 side view centered below it, 1 front view at the lower left, and 1 boxed detail inset at the lower right showing the shell rib pattern. Use white contour lines and dashed internal rib lines.

Technical text: Add a specification block at upper left with exactly 4 rows: TYPE: {argument name="type label" default="Small Shell Cake"}, ORIGIN: {argument name="origin label" default="France"}, SHAPE: Ovoid / Shell, UNITS: Millimeters.

Measurements: Add dimension lines and numeric labels matching a blueprint: top view width 75 and height 45, side view height 22, front view width 45 and height 22. Add a bottom-left scale ruler labeled SCALE with marks 0, 10, 20, 30, 40, 50 mm.

Additional elements: Add the inset label “DETAIL A” and “SHELL RIB PATTERN”. Add a notes box at the bottom right with exactly 2 notes: “1. All dimensions are approximate.” and “2. Typical madeleine proportions may vary.”

Style constraints: Do not render the pastry as a realistic food photo; convert it fully into a white-on-blue technical schematic. Keep the blueprint crisp, precise, lightly hand-drawn at the outline edges, with no extra objects, no shadows, and no watermark.

Goal: Create a glossy YouTube thumbnail-style image for an ASMR streamer niche, combining a realistic cozy creator photo with large Japanese title graphics and neon social-media icons.

Canvas: Horizontal 16:9 thumbnail, 1152×768 style composition, cinematic shallow depth of field, cool blue and purple lighting with soft pink highlights.

Main subject: A young female ASMR YouTuber sits on the right half of the frame in a cozy bedroom or streaming room, wearing an off-shoulder lavender knit sweater and black in-ear monitors. Her face is intentionally covered by a large flat rectangular privacy block in muted brown, centered over the face from forehead to chin. She holds a large black condenser microphone with silver mesh grille close to her mouth, mounted in a black shock mount on a stand. Her visible hands have neat pale nails. The mood is intimate, calm, and whisper-stream themed.

Background: Soft-focus bedroom/studio with curtains, shelves, small plants, fairy lights, and a bed or sofa. Use blue-purple ambient lighting throughout, with pink bokeh lights in the center background. Add a glowing mushroom-shaped table lamp at the bottom right and white-and-lavender headphones partially visible beside it.

Layout: Put the creator and microphone on the right, and place a large translucent rounded rectangle title panel across the lower-left to center-left. The panel has a blue-to-purple gradient glassmorphism look. On the far left of the panel, include one large circular gradient badge containing the number “1.” in bold white. To the right of the badge, place the main headline in bold rounded white lettering: “ASMR・配信・” on the first line and “YouTuber” on the second line. Beneath it, add a smaller white pill-shaped label containing “ささやき・会話・配信向け” in blue-to-pink text.

Visible text: Use exactly 3 text blocks: 1) “1.” inside the circle badge, 2) main headline “{argument name="headline text" default="ASMR・配信・YouTuber"}”, 3) subtitle “{argument name="subtitle text" default="ささやき・会話・配信向け"}”. Keep all text crisp, readable, and faithful to the Japanese characters.

Decorative elements: Include exactly 7 neon/social graphic elements around the scene: 1) glowing cloud outline in the upper left, 2) white neon microphone icon below the cloud, 3) white neon chat bubble with three dots beside it, 4) glowing play-button outline in the upper right, 5) audio waveform line along the lower left, 6) one white music-note icon near the bottom center, 7) multiple small sparkle stars scattered near the icons and right edge as a single sparkle cluster style. Use purple, white, and pink neon glow.

Visual style: Realistic photo base with overlaid clean vector graphics, modern creator-economy thumbnail design, soft bloom, high contrast, glossy gradients, rounded typography, dreamy ASMR atmosphere, no watermark, no extra logos, no extra text. Use {argument name="main color palette" default="lavender, blue, purple, and soft pink"} lighting and emphasize {argument name="content niche" default="ASMR streaming and YouTuber prompts"}.

Create a vibrant Japanese marketing course thumbnail for {argument name="campaign topic" default="CM・広告・商品訴求"}. Use a 16:9 horizontal canvas, 1184×768 style, with a glossy purple-blue neon aesthetic and soft depth-of-field office lighting. Background: a modern cosmetics marketing workspace with a blurred large monitor showing a campaign dashboard; visible English text on the monitor includes “NEW CAMPAIGN” and small chart/product panels. On the right, place a young woman marketer in a lavender blouse, seated at a desk with one hand near her chin holding a pen; her face is intentionally hidden by a large opaque matte brown rectangle. Desk details include exactly 5 main objects: an open laptop on the left, a disposable coffee cup in front, printed marketing sheets with colorful layout thumbnails, cosmetics bottles on the right, and a small sign reading “NEW ARRIVAL 30% OFF LIMITED CAMPAIGN.” Overlay a large rounded gradient banner across the lower-left to center, blue-to-purple with glow and transparency. On the far left of the banner, put a large circular badge with the white text “2.”. Main Japanese headline on the banner: “CM・広告・商品訴求” in bold white rounded sans-serif, with “商品訴求” on the second line. Beneath it, add a white pill-shaped label with blue and pink text: “商品訴求・販促・告知向け”. Add exactly 3 large neon line icons: a cyan megaphone in the upper left, a pink shopping bag below it, and a purple play button inside a rounded rectangle on the upper right. Add exactly 8 small glowing sparkle accents scattered around the composition, plus a small pink upward bar-chart icon near the bottom center. Overall style: Japanese YouTube thumbnail, polished digital advertising design, luminous gradients, cinematic blur, clean readable typography, no watermark, no extra text beyond the specified labels.

Goal: Create a vibrant Japanese travel-video title card for a short-form prompt collection, combining a cinematic spring travel photo background with glossy neon UI graphics.

Canvas: Horizontal 16:9 image, 1152×768 feel, high-resolution, bright pastel color grading, soft bloom and lens flare.

Background: A young woman traveler stands on the right side at a train platform beside a modern train, viewed from behind/three-quarter back, face intentionally obscured by a plain skin-tone rectangular blur. She has long wavy dark brown hair, wears a light beige trench coat, carries a tan shoulder bag, and holds the handle of a small rolling suitcase near the bottom center. The setting is a dreamy cherry-blossom station in spring, with pink sakura trees filling the left and upper background, scattered falling petals, warm sunlight, and shallow depth of field.

Main graphic layout: Place a large rounded gradient banner across the left-center, extending toward the woman. The banner uses a blue-to-purple-to-pink translucent glossy gradient with a large circular badge on the far left. Inside the circle, display a huge white number {argument name="section number" default="4."}. To the right of the circle, set bold rounded Japanese headline text in white: {argument name="headline text" default="旅行・季節・青春シネマ"}. Below it, add a smaller white rounded pill label with purple-pink text: {argument name="subtitle text" default="情緒・旅感・短編向け"}.

Visible text count: exactly 3 text elements: 1 large number badge reading "4.", 1 main Japanese headline reading "旅行・季節・青春シネマ", and 1 subtitle pill reading "情緒・旅感・短編向け".

Neon icon overlay count: exactly 9 glowing white-pink line icons and decorative elements around the banner and right side: 1 suitcase icon at upper left, 1 cherry blossom icon near upper center-left, 1 camera icon above the banner, 1 play button icon inside a rounded rectangle on the upper right, 1 clapperboard icon on the mid-right, 1 signpost arrow icon at lower left, 1 dotted travel path sweeping along the lower left-to-center, 1 map pin near the end of the dotted path, and 2 small sparkle/star accents near the top center and right side. Keep all icons semi-transparent, neon outlined, and softly glowing.

Style: Japanese social-media thumbnail design, cute travel aesthetic, glossy gradient panels, soft pink and lavender atmosphere, cinematic bokeh, dreamy sakura petals, clean rounded typography, bright white text with subtle glow, polished promotional-card composition.

Constraints: Keep the woman on the right and the graphic title block on the left, do not add extra text, do not add logos or watermarks, keep the face covered by a simple rectangular blur, and maintain a cheerful spring travel mood.

Create a square 1:1 close-up food advertisement photo of a cheesy quesadilla, shot in warm golden light with shallow depth of field. The foreground shows two stacked triangular halves of a toasted flour tortilla quesadilla on a dark rustic plate, cut open so the filling is highly visible: browned spiced ground beef, glossy sauce, and abundant melted yellow-white cheese stretching in long strands between the top and bottom pieces and dripping over the edges. The tortilla surface should be crisp with scattered golden-brown char marks and soft blistering. Use a dark wooden table background, softly blurred, with a small out-of-focus bowl of greens and cherry tomatoes in the upper right. Overlay energetic Japanese handwritten brush typography: exactly 3 text elements, consisting of 1 small orange handwritten line at the upper left reading {argument name="top handwritten text" default="チーズたっぷり"} with a curved underline flourish, 1 large bold white brush-script headline across the upper center reading {argument name="main headline text" default="ケサディーヤ"}, and 1 smaller white handwritten note at the lower left reading {argument name="bottom handwritten text" default="とろ〜り チーズ"}. Add exactly 4 orange doodle accents: a thick orange brush underline sweeping beneath the main headline, two short orange emphasis marks near the lower left edge, three short orange rays near the upper right of the headline, and a small orange heart beside the lower-left note. Make the typography look like casual hand-painted Japanese marker lettering, slightly imperfect and organic. Keep the food hyper-realistic, appetizing, high contrast, with saturated cheese yellows, rich beef browns, and a cozy restaurant-menu feel. No people, no logos, no watermarks, no extra text.

Create a high-quality vertical anime illustration of {argument name="character name" default="Reina"}, a Japanese high school girl taking a close-up arm-extended selfie inside a sunlit classroom. She has very long flowing {argument name="hair color" default="dark forest green hair with pale mint-green tips"}, glossy layered strands, and one hand raised near her temple as if adjusting her hair. Her face is intentionally covered by a large centered opaque square censor block with a muted gray-to-beige color, hiding all facial features from forehead to chin. She wears a neat navy school uniform blazer with gold buttons, a crisp white shirt, and a large {argument name="bow color" default="deep red ribbon bow"} at the collar, with a pleated navy skirt partially visible. The camera angle is a dynamic wide-angle selfie perspective from slightly above and in front, with her extended left arm dominating the lower-left foreground and her torso centered. Background: a bright empty classroom with wooden desks and chairs, a green chalkboard on the left, tall windows on the right, warm morning sunlight streaming in, soft bloom, dust motes, and gentle depth of field. Style: polished modern anime key visual, detailed hair highlights, clean line art, soft painterly shading, warm cinematic lighting, vertical 3:4 composition, no text, no watermark, no extra characters.

Create a realistic candid smartphone photo of a young woman at an outdoor city event, shot from a close slightly high-angle perspective with a wide-angle lens. She is leaning playfully toward the camera, making a sideways peace sign beside her face with her right hand, while her left arm hangs down. Her face is intentionally covered by a flat opaque square censor block in a muted skin-tone color. She has a high messy ponytail with {argument name="hair color" default="pink peach hair with a small teal-blue streak"}, loose wispy strands, small earrings, rings, and glossy decorated nails. She wears a fitted dark indigo denim halter overall jumpsuit with visible seams, structured bust cups, thick straps, and a snug waist, giving an energetic street-fashion look. The background is an outdoor festival or expo entrance on asphalt, with white peaked event tents, blue crowd-control railings, scattered pedestrians slightly motion-blurred, tall modern apartment towers, cloudy daylight sky, and shallow depth of field. Use natural lighting, realistic skin texture, casual Japanese urban atmosphere, vertical 2:3 composition, candid social-media photo style, no text, no watermark.