GPT Image 2 prompts

Goal: Create a vertical old-magazine narrative poster about a past blind-date relationship story, in a gritty Chinese editorial zine style, using one young woman as the only person shown.

Canvas: Portrait 2:3 poster, aged off-white paper background with gray stains, black ink smears, torn-paper edges, photocopy grain, rough brush textures, red distressed accent marks, and a slightly messy collage layout.

Subject details: Show one East Asian young woman with {argument name="hair style" default="long straight black hair"}, pale skin, soft makeup, and a quiet distant expression. She wears {argument name="outfit" default="a black off-shoulder knit top"} in every photo. Keep her face consistent across all appearances.

Layout: Use exactly 4 black-and-white/sepia-tinted photo panels of the same woman: 1 large main portrait near the upper center-right where she holds a phone and looks sideways with a cold, reserved expression; 1 small photo at bottom-left where she drinks through a straw while looking down; 1 small tilted photo at mid-right where she rests her chin on her hand and looks down; 1 small tilted photo near the bottom-right where she looks at a phone with a faint restrained smile. Photos should have torn borders, uneven crop edges, taped scrapbook placement, and dark grunge overlays.

Typography and text: Make the poster text-heavy like a Chinese emotional magazine spread. Use bold Chinese headline typography with black body copy and deep red emphasis. Place a huge vertical title on the right reading {argument name="main title" default="相亲往事"}. Add small English subtitle beside it: “PAST DATING STORIES”. Add two slim vertical Chinese side notes near the title: “有些人错过了是遗憾” and “有些人没开始反而是幸运”. On the left and lower center, arrange multiple Chinese story paragraphs in compact columns, resembling an autobiographical dating essay. Include one large bold pull quote near the lower right reading {argument name="pull quote" default="感情里最累的，不是明确的拒绝，而是她冷着你，却有人一直让你再主动一点。"}. Include another bold paragraph on the left with red final line reading {argument name="left emphasis text" default="后来聊天不冷不热，约她出来也总说忙，我慢慢觉得不合适，就不再主动联系了。"}. Add a black rectangular caption block near the lower middle with white Chinese text, plus a red quoted sentence below it reading {argument name="accusation quote" default="你是不是在亲相亲没人要，没人相中你，所以又来找我聊？"}.

Visual style: Melancholy, nostalgic, urban, emotionally restrained, like a 1990s distressed magazine page mixed with modern social-media storytelling. High contrast monochrome portraits, muted red ink, rough halftone texture, torn collage, imperfect alignment, and worn printed-paper grain.

Constraints: Use only the same woman, no other people. Keep exactly 4 photo panels. Keep the overall composition dense but readable. Do not add logos, watermarks, QR codes, or extra decorative characters.

Goal: Create a wide single-panel chibi anime comic illustration about a prison breakout scene, split visually into two labeled areas: a dark jail cell on the left and a brighter guard corridor on the right.

Canvas: 16:9 horizontal image, polished digital anime style, cute chibi proportions, clean line art, soft cel shading, dramatic lighting from the wall impact, high detail but readable composition.

Layout: The left two-thirds show the prison cell interior with blue-gray stone walls, barred doors, a metal bed, and a toilet. The right third shows the guard corridor outside the bars, lit with warmer beige light. Use a vertical boundary created by the cell bars and wall. Add a large cracked hole being smashed into the stone wall near the center-right of the cell, with debris, dust, and a bright orange impact flash.

Text content: Include exactly 6 visible Japanese text elements: 1 top-left white label box reading {argument name="left area label" default="牢獄"}; 1 top-right white label box reading {argument name="right area label" default="看守通路"}; 1 small label near the toilet reading {argument name="toilet label" default="トイレ"}; 1 bottom-right label reading {argument name="bottom label" default="サムズアップ　看守通路"}; 1 speech bubble above the pink-haired guard reading {argument name="first speech bubble" default="怪しいアイドルね？"}; 1 speech bubble above the lavender-haired guard reading {argument name="second speech bubble" default="怪しすぎる元アイドルですね。"}. Add small stylized sleep sound effects around the sleeping prisoner, including Japanese-style blue “zzz” marks and a sleepy ellipsis.

Characters: Include exactly 5 chibi female characters. Character 1: a blonde idol prisoner in the center of the cell, energetic and smiling, wearing a black-and-gold idol combat outfit with star accessories and thigh-high boots, mid-kick with one leg extended into the stone wall, creating the explosion and cracks. Character 2: a pink-haired twin-tail guard outside the bars near the lower right center, wearing a navy-and-white uniform with a small black cap, giving a thumbs-up while watching the breakout. Character 3: a red/pink-haired guard in a dark military-style uniform and red-star cap, standing to the right with a suspicious worried expression, sweat drops beside her face. Character 4: a lavender-haired bespectacled guard in a white sailor-like uniform and white cap, standing at the far right with a calm skeptical expression. Character 5: a sleeping prisoner on the upper-left metal bed, pale pink hair, wearing a dark outfit and cap, curled on her side with closed eyes and blue sleep symbols.

Props and environment: Include exactly 2 major cell fixtures: 1 metal bed at the upper left and 1 white toilet at the lower left. The toilet has a cloth doll or plush idol with X-shaped eyes sitting on it. Add metal bars in the foreground and background, stone block walls, cracked masonry, scattered rubble, and a small barred window high on the back wall.

Visual style: Cute chibi anime, expressive faces, slight isometric perspective, comic panel composition, detailed prison background, contrast between cool blue prison lighting and warm corridor lighting, no photorealism.

Constraints: Keep the scene as one continuous illustration, not a multi-panel grid. Preserve the exact count of 5 characters, 2 main fixtures, 6 Japanese text elements, and 2 speech bubbles. Do not add extra characters, extra panels, watermarks, or English replacement text.

Create a cinematic fashion editorial photograph in a cold urban street at dusk: a young adult woman with {argument name="hair color" default="long wavy blonde hair with darker roots"} sits curled on a worn olive-tan vintage sofa placed directly in the middle of a wet city road. She wears an oversized {argument name="coat style" default="cream faux-fur coat"}, black fishnet tights, and chunky black lace-up combat boots; her legs are crossed, arms folded close to her body, and she stares intensely into the camera with a moody, slightly defiant expression. The sofa has exactly four visible cushions: one long back cushion, one large seat cushion, and two loose side pillows, with damp stains and scuffed fabric. Surround the scene with gritty downtown architecture: tall brick and stone buildings on both sides, arched windows on the right, parked cars along the curb on the left, exactly two dark mailboxes/trash receptacles behind the sofa, one yellow school bus in the distant center-left, and small piles of melting snow along the sidewalks and curb. The street surface should be wet and reflective, with scattered slush, muted winter colors, and a cold blue-gray atmosphere. Use shallow depth of field with the model and sofa sharp, background softly blurred, wide-angle street perspective, low eye-level camera, realistic film grain, subtle lens flare or blown streetlight glow in the upper left, slight analog scratches, high-end fashion photography styling, natural overcast dusk light, dramatic contrast, no text, no watermark.

Create a highly detailed anime illustration of one cat-eared girl named {argument name="character name" default="Miu"} lying on her stomach in a graceful side pose, propping her chin on one hand and gazing softly toward the viewer with bright {argument name="eye color" default="cyan blue"} eyes. She has very long flowing {argument name="hair color" default="silver white"} hair with curled strands spread across the floor, straight bangs, two large fluffy white cat ears, one burgundy flower hair ornament with a gold center, one small burgundy X-shaped hairpin, small earrings, and one matching flower choker. Dress her in an ornate {argument name="kimono color" default="deep burgundy"} gothic-lolita kimono dress with wide sleeves, intricate white floral embroidery, lace trim, layered ruffled skirt, patterned bow at the back, black inner collar details, and delicate translucent lace accents. Show her bare legs bent upward behind her, with both feet visible near the upper right. The setting is a pale decorative Japanese floral backdrop with cherry blossom branches, scattered white petals and flowers, a burgundy crescent moon motif, faint checkered texture, an ornate thin lace-like rectangular border, and a glossy reflective floor that mirrors her sleeve, hair, and blossoms. Use a soft pastel palette of burgundy, white, pale pink, and gray, with clean polished line art, luminous highlights, delicate shading, and an elegant dreamy AI-anime art style. Include exactly these visible repeated elements: one girl, two cat ears, two raised bare feet, three main burgundy accessories on her head/neck area, one crescent moon symbol, many cherry blossom branches and scattered petals. Add a small handwritten signature reading {argument name="signature text" default="ausa days"} with a tiny paw mark in the bottom right corner. No extra characters, no modern objects, no watermark besides the small signature.

Create a vertical black-and-white manga illustration in highly detailed ink line art and screentone shading. Scene: a quiet moonlit traditional Japanese wooden veranda at night, open to a garden. In the foreground, place exactly one round wooden washtub filled with water and exactly 14 small square ice cubes, with exactly one normal-sized hamster floating on its back in the cold bath, relaxed and tiny compared with the tub. Beside the tub, show exactly one androgynous beautiful young adult in a loose summer kimono/yukata, {argument name="character name" default="unnamed pale-haired young man"}, with {argument name="hair color" default="very light silver-white hair"}, long messy bangs, and one thick side braid over the shoulder. The character kneels on the veranda, leaning forward with a melancholy, gentle expression, holding exactly one round paper uchiwa fan decorated with simple floral and grass motifs, as if fanning the hamster. Clothing: flowing pale yukata with subtle flower patterns, dark obi sash, wide sleeves, traditional folds and fine fabric texture. Background: traditional wooden beams, sliding shoji doors, interior shadows, an outdoor garden with leafy shrubs and hydrangea-like flowers, exactly one stone lantern, exactly one crescent moon in a starry black sky. Add small water drips and puddles on the wooden floor near the tub, dramatic moonlit contrast, dense crosshatching, delicate hair strands, cinematic composition, intimate summer-night atmosphere. Use {argument name="art style" default="Japanese manga engraving style"}; keep the hamster realistic common-pet size, not giant, and avoid color, text, speech bubbles, modern objects, or extra animals.

Create a vertical black-and-white manga illustration in extremely detailed pen-and-ink crosshatching, set on a quiet traditional Japanese wooden veranda at night. A mysterious {argument name="character name" default="young woman"} with long messy {argument name="hair color" default="black hair"} tied in a loose bun kneels beside a round wooden washtub, wearing layered flowing {argument name="outfit" default="pale kimono robes with dark inner layers, cords, tassels, and loose sleeves"}. She leans forward with a gentle, slightly unreadable expression, holding a flat round paper fan painted with grasses and flowers, fanning a very large fluffy {argument name="animal" default="hamster"} relaxing belly-up in cold water. The hamster looks blissful with closed eyes, tiny paws raised, and a small folded towel with dotted pattern resting on its head. The tub is filled with water and exactly 24 visible translucent ice cubes: 8 clustered along the front rim, 6 floating near the hamster’s left side, 5 around the right side, and 5 behind the hamster near the back of the tub. The scene includes a wooden deck with strong plank texture, sliding glass doors standing open, a dark garden outside, leafy silhouettes, a stone lantern glowing softly on the left, and the interior of a study on the right with bookshelves, stacked books, a desk lamp, and a vertical hanging calligraphy scroll reading {argument name="calligraphy text" default="観書養志"}. Use dramatic moonlit lighting with deep shadows, a bright diagonal beam falling across the sliding door, soft glow from the lantern and desk lamp, intricate linework, dense textures, melancholic summer-night atmosphere, no color, no modern objects, no speech bubbles, no captions.

Create a dramatic square-format anime portrait of {argument name="character name" default="a mysterious young woman"} in a pitch-black void, shown from the chest up in a three-quarter pose facing right while looking slightly back toward the viewer. She has long flowing {argument name="hair color" default="platinum blonde"} hair with heavy bangs, glossy strands, and a glowing rim-lit section down the left side. She wears a wide-brimmed dark hat with oversized rabbit-ear-like extensions or sharp folded panels, plus a dark floral-patterned outfit with large pale teal flowers and soft pink accents. Use extreme low-key lighting: most of the image is swallowed by black shadow, with intense cyan-blue backlight from the left, subtle magenta-pink illumination on the face and lips, and small sharp highlights on the eyes, hair, hat edge, shoulder, and fabric. The expression should be calm, distant, and slightly melancholic, with parted lips and large detailed eyes. Add one small cyan triangular shard of light floating in the darkness to the right of her face. Style the image as polished high-contrast Japanese anime key art with painterly digital brushwork, cinematic noir lighting, luminous edge glow, deep shadows, soft bloom, elegant composition, and no visible text or watermark.

Goal: Create a clean black-and-white Japanese manga page about a shy high-school basketball boy and a supportive girl, using delicate screentone shading, hand-inked linework, cinematic paneling, and readable vertical Japanese speech bubbles.

Canvas: A single portrait manga page on off-white paper with a rounded light-gray page border. Use a 2-column by 3-row grid with exactly 6 rectangular panels separated by thick white gutters. No color except grayscale ink, screentones, and soft highlights.

Characters: The main boy is {argument name="male lead" default="a slim teenage Japanese basketball player with messy short black hair, jersey number 7, athletic shorts, sneakers, and a shy earnest expression"}. The main girl is {argument name="female lead" default="a teenage Japanese schoolgirl with straight shoulder-length black hair and bangs, sailor-style school uniform, warm smile, sometimes holding a book or towel"}.

Panel layout and content: Panel 1, top left: indoor school gym, basketball hoop close to camera, polished floor reflections, the boy in jersey number 7 jumps for a layup or shot, ball near the rim, dynamic upward motion, one vertical speech bubble. Panel 2, top right: outdoors by a chain-link fence at school, the boy stands with a backpack facing the girl, she holds books against her chest, calm encouraging mood, one vertical speech bubble near the girl. Panel 3, middle left: dramatic close-up under the basketball hoop, ball arcing toward the net, swirling speed lines fill the background, boy looks up intensely and nervously, one vertical speech bubble. Panel 4, middle right: bright school corridor or gym hallway with large windows, the smiling girl hands the boy a towel while he holds a drink bottle and looks flustered, exactly two vertical speech bubbles. Panel 5, bottom left: inside the gym, boy crouches alone tying or adjusting his shoe, looking determined and embarrassed, bleachers behind him, a soft glowing thought-image portrait of the girl appears above and behind him, one vertical speech bubble. Panel 6, bottom right: school courtyard in spring with cherry blossoms and petals falling, school building in the background, boy in basketball uniform and girl in school uniform face each other with shy smiles, exactly two vertical speech bubbles.

Text content: Use the following Japanese dialogue in the 8 speech bubbles, arranged by panel order: {argument name="Japanese dialogue" default="絶対もっと上手くなる！ / いっしょに頑張ってるね。 / やばい…なんかすごい意識しすぎ… / おつかれさま。 / あ、ありがとう…？ / もっと強くなって、ちゃんと想いを伝えたい。 / 好きです。付き合ってください。 / うん、よろしくね"}. The lettering should be printed manga-style, vertical where appropriate, crisp and legible.

Visual style: Authentic shonen romance manga, monochrome ink, fine hatching, screentone shadows, expressive eyes, soft romantic lighting, detailed gym architecture, chain-link fence texture, window reflections, cherry blossom petals, dynamic speed lines in action panels, gentle emotional pacing.

Constraints: Exactly 6 panels, exactly 2 main characters, boy jersey number must be {argument name="jersey number" default="7"}, keep all speech bubbles inside panel borders, no extra captions, no logos, no watermark, no color.

Using REFERENCE_0 as the base image, regenerate the same scene while preserving its steampunk gothic city atmosphere, cinematic lighting, color palette, architectural density, elevated railway, train, airships, fog, smoke, and overall painterly realism. Change only the camera angle: move the camera much higher and slightly farther back, creating a wider bird’s-eye / high-angle view that reveals more of the city below, more of the viaduct structure, and more surrounding cathedral-like towers. Keep the train on the elevated track in the lower-right foreground, but make it smaller in frame due to the raised viewpoint. Maintain the same moody dusk haze, warm window lights, and epic scale; avoid modern elements, clean sci-fi styling, or a flat top-down view.

Goal: Create a richly detailed fantasy anime illustration of a desert starlight cafe during a Journey to the West-inspired stopover, with a cozy magical atmosphere under a vast Milky Way sky.

Canvas: Vertical 4:5 illustration, high resolution, ultra-detailed, warm lantern light contrasted with deep blue and violet night sky. The scene is set at twilight-night in a desert oasis-like roadside cafe.

Main setting: A vintage propeller airplane has been converted into a small open-air cafe. The fuselage forms the shop body on the left, with the long wing extending across the upper right like an awning. The airplane is weathered cream metal with rivets, vines, flowers, and small plants growing over it. Hang glowing lanterns and string lights beneath the wing. Add a flowing translucent golden cloth draped from the wing and blowing gently in the wind. In the background show mesas, cacti, desert shrubs, warm sunset glow at the horizon, and an immense star-filled sky with a bright Milky Way band and one shooting star.

Characters and exact count: Include 6 character figures plus 1 cat. 1) {argument name="monk character" default="a beautiful Sanzo monk"}: seated elegantly on a tall wooden stool at the cafe counter, wearing ornate white and red-gold Buddhist-inspired robes and a jeweled crown, long pale hair, calm expression, holding a coffee cup under the stars. 2) {argument name="monkey warrior" default="Sun Wukong"}: crouching in the lower left foreground in yellow and red warrior clothing, golden circlet, mischievous excited expression, reaching toward a glowing star. 3) A gentle gray-haired cafe attendant girl behind the counter, wearing a white blouse and beige apron with star motifs, holding a coffee pot, soft anime face, flowers in her hair. 4) A sleepy pig-like companion in the rear right, reclining on cushions under a small shade structure. 5) A bearded river-sand monk companion beside him, also asleep. 6) A white dragon horse resting at the far right near them. 7) A white-and-orange cat in the center foreground standing on hind legs, holding up a bright glowing yellow star that fascinates the monkey warrior.

Cafe objects and exact counts: Show 3 visible cafe sign/menu boards: one hanging wooden sign on the right reading {argument name="hanging sign text" default="星空咖啡 / Starry Desert COFFEE"}, one blackboard menu behind the counter reading “Starlight Blend” with drink names such as drip coffee, latte, moon latte, star cookie, and galaxy soda, and one freestanding chalkboard sign in the lower left reading “Starlight Blend” with a small coffee cup drawing. Include many coffee tools on the counter: glass dripper, kettle, bottles, cups, jars, saucers, small pastries, and star-shaped sweets. Include exactly 5 prominent blue glass bottles in the lower right foreground, some containing tiny glowing stars.

Foreground details: Stone-paved cafe terrace with scattered tiny star lights, flowers, potted plants, desert blossoms, wooden crates with a star emblem, a lantern glowing near the lower left, and small handwritten tags. Use abundant tiny decorative details: vines, yellow and blue flowers, jars of light, hanging lamps, coffee accessories, and star motifs.

Visual style: Painterly Japanese anime fantasy, extremely intricate background art, luminous magical realism, soft character rendering, cinematic composition, vibrant saturated colors, warm gold highlights, cool indigo sky, cozy whimsical travel mood, highly detailed 5k illustration quality.

Constraints: Preserve the desert starlight cafe theme, keep the airplane cafe clearly recognizable, keep the composition crowded but readable, avoid modern city elements, avoid photorealism, avoid extra characters beyond the 6 figures and 1 cat.

Create a photorealistic fashion portrait of a young {argument name="woman description" default="blonde woman with a shoulder-length bob"} standing on a bridge or riverside railing in Paris at golden hour. She faces the camera with a soft confident smile, wearing thin round eyeglasses, a crisp white button-down blouse with a slightly open collar, a glossy black patent leather trench coat, a high-waisted glossy black patent leather mini skirt, and sheer black tights. Her left arm rests casually on a metal railing while her right arm hangs naturally, upper-body three-quarter framing from mid-thigh upward. The background shows the Seine river with shimmering sunset reflections, classic Haussmann buildings on the left, an arched bridge, and the Eiffel Tower softly blurred in the distance on the right. Use warm sunset backlighting, realistic skin texture, shallow depth of field, cinematic bokeh, high-fashion editorial styling, sharp focus on the face and shiny black vinyl textures, natural proportions, no text, no watermark.

Goal: Create a high-resolution editorial angle reference sheet featuring the same fashion model consistently across exactly 9 panels.

Canvas: Vertical 4:5 aspect ratio, arranged as a perfect 3x3 grid with thin white dividers between panels. Use a neutral gray studio backdrop in every panel, soft editorial lighting, realistic photography, crisp detail, and consistent color grading.

Subject details: A smiling young adult woman with fair skin, long wavy {argument name="hair color" default="platinum blonde"} hair, blue eyes, polished makeup, and bright red lipstick. Maintain identical facial identity, age, hair texture, skin tone, and proportions across all panels. Wardrobe is fixed in every panel: a deep {argument name="suit color" default="red"} tailored blazer and matching trousers, white collared shirt, long black necktie, and black shoes. Confident editorial fashion styling.

Layout and required panels: Include exactly 9 separate portraits, each with a bold white numbered caption centered near the bottom of its panel:
1. “1. Front Medium” — medium front-facing portrait, smiling, torso visible.
2. “2. Close-Up” — tight beauty close-up of the same face, smiling, hair framing the face.
3. “3. Three-Quarter” — three-quarter view from the waist up, turned slightly to one side.
4. “4. Profile” — clean side profile portrait, facing left.
5. “5. Low Angle Full” — full-body low-angle fashion pose, hands in pockets, legs apart.
6. “6. High Angle” — high-angle medium portrait, looking up toward camera.
7. “7. Over Shoulder” — over-the-shoulder view with the model turned away but looking back and smiling.
8. “8. Dutch Tilt” — tilted-camera medium fashion portrait, hands in pockets, playful confident pose.
9. “9. Mid Walk” — full-body walking pose, stepping forward, smiling at camera.

Visual style: Editorial studio photography, realistic lens rendering, professional fashion reference sheet, balanced exposure, shallow-to-moderate depth of field depending on shot type. Keep the background plain and consistent.

Constraints: Use exactly the 3x3 grid and exactly the 9 labeled panels listed above. Keep the same model identity and outfit in all panels. Do not add extra panels, props, logos, watermarks, or decorative text beyond the panel captions.

Create a surreal forced-perspective ultra-realistic photograph of the same young woman appearing at two different scales in the middle of a modern downtown city street during golden hour. The giant version of {argument name="character name" default="the woman"} is crouching low on the right side of the empty roadway between tall glass skyscrapers, smiling gently and looking down as she reaches one hand toward a tiny full-body version of herself standing on the left side of the street. The tiny version wears the same outfit and raises one arm upward, touching or almost touching the giant woman’s fingertip, creating a playful miniature-meets-giant illusion. She has {argument name="hair color" default="dark brown hair"} tied in a high ponytail with loose strands, natural makeup, hoop earrings, and a soft friendly expression. Outfit: loose {argument name="shirt color" default="black"} oversized t-shirt, beige cargo pants with side pockets, chunky white sneakers, and a light-colored watch on the giant figure’s wrist. Use realistic anatomy and proportions at both scales, one giant crouching figure and one miniature standing figure only. Set the scene on a wide asphalt avenue with warm sunlight glowing directly down the canyon of skyscrapers, long shadows, reflective glass buildings, a few blurred cars and pedestrians far in the background, traffic lights, and cinematic depth of field. Camera is low to the road at street level, vertical portrait composition, dramatic golden-hour backlight, warm highlights on the pavement, sharp focus on both figures, photorealistic texture, ultra-detailed clothing folds, natural skin, no text, no watermark, no extra duplicate people.

Create a photorealistic close-up product-style image of a handcrafted wooden travel chessboard on a walnut tabletop. The scene shows exactly 1 folding chessboard made from 2 rectangular wooden halves separated by a narrow central hinge groove, with a checker pattern of exactly 64 raised square tiles in alternating pale maple and dark walnut wood. Add exactly 2 loose accessory pieces beside it: 1 long dark walnut side rail resting parallel to the right edge of the board and 1 flat pale maple wooden strip lying farther right. Include exactly 1 small brass-colored latch or clasp visible on the near front edge of the board. Use warm natural window light from the upper left, shallow depth of field, realistic wood grain, beveled edges, soft shadows, and a cozy workshop or tabletop atmosphere. Camera angle: three-quarter overhead view, slightly tilted, focused on the board surface. Make the board appear like {argument name="product type" default="a foldable travel chessboard"} made of {argument name="dark wood" default="walnut"} and {argument name="light wood" default="maple"}, placed on {argument name="surface" default="a walnut table"}. No chess pieces, no people, no text, no watermark.

Create a cinematic close-up of an ancient Turkic/Oghuz warrior’s ornate bronze-gold armor, focusing on the chest plate rather than the face. The breastplate features a large raised snarling wolf head emblem at the center, with pointed ears, sharp eyes, open mouth, visible teeth, and sculpted fur details. Surround the wolf with embossed swirling ornamental patterns, rivets, dark leather straps, and weathered metal plates. Show both shoulder pauldrons partially in frame, decorated with engraved motifs and layered armor segments. Use {argument name="armor material" default="aged bronze and darkened steel"} with tarnish, scratches, patina, and realistic battle-worn texture. The warrior is mostly cropped from the neck down, with only a hint of long dark reddish hair above the collar. Set the scene in a blurred medieval battlefield or fortress environment at sunset, with warm golden rim light from the right and deep shadows on the left. Use shallow depth of field, dramatic cinematic lighting, anamorphic film look, ultra-realistic detail, 16:9 widescreen composition, no text, no logos, no modern elements.

Create a photorealistic vertical outdoor portrait of a beautiful young East Asian woman sitting on a wooden park bench in a lush green park during warm late-afternoon sunlight. She has long straight {argument name="hair color" default="black"} hair with soft bangs and a small hair clip on one side, clear fair skin, natural makeup, and a gentle shy smile while looking slightly off-camera. She wears a fitted {argument name="dress color" default="white"} off-shoulder lace mini dress with short lace sleeves, a corset-style lace-up front with thin ribbons, delicate floral lace texture, and a scalloped lace hem. Pose her seated sideways on the bench with both hands resting neatly on her knees, relaxed shoulders, elegant posture, and legs angled toward the foreground. The setting includes exactly 1 wooden bench, 1 paved green walking path curving into the background, 1 green trash bin, 1 wooden park signpost with small unreadable signs, several distant pedestrians softly out of focus, dense trees and bushes, dappled shadows, and golden sunlight filtering through leaves. Use a realistic smartphone photo aesthetic, natural skin texture, shallow depth of field, crisp subject focus, soft background bokeh, 3/4 portrait framing from knees to head, vertical 4:5 composition, no visible watermark, no added text, and no exaggerated fantasy styling.

Create a dark fantasy cinematic key art scene in an ultra-wide 16:9 frame. In the foreground left, show one lone armored knight, back turned toward the viewer, wearing battered blackened plate armor and a torn deep {argument name="cape color" default="blue"} cloak with gold frayed trim, gripping a long straight sword planted point-down in the wet stone ground. In the midground and center-right, show one colossal undead cathedral-like giant towering over the knight, with a helmet-like vertical slitted glowing face, exposed ribcage armor, massive hunched shoulders, long dangling arms, thick legs, and heavy strands of moss, roots, chains, and tarnished gold filigree hanging from its body. The scene contains exactly two humanoid figures: the small knight and the enormous giant. Surround them with exactly three prominent ruined gothic structures: a tall tilted spire on the far left edge, a broken arched bridge behind the knight in the middle-left background, and a dark ruined cathedral arch tower on the far right. Set the environment in a ruined medieval city or cathedral courtyard under a stormy night sky, with low fog, ash particles, scattered tiny torch flames, wet reflective stone, and a red-black storm glow on the left fading into cold blue moonlight on the right. Use {argument name="visual style" default="grimdark photorealistic fantasy concept art"}, extremely detailed armor, monumental scale, high contrast chiaroscuro lighting, dramatic low camera angle, atmospheric haze, sharp focal detail on the knight and giant, no text, no logos, no modern objects.

Goal: Create a vertical scrapbook-style travel guide poster for {argument name="destination city" default="Chengdu"}, combining cute hand-drawn decoration, taped paper scraps, Polaroid photos, Chinese calligraphy titles, and food/travel stickers.

Canvas: Tall portrait poster, 9:16 ratio, soft sky-blue paper-texture background with white torn-cloud shapes near the top and throughout. Use a warm, playful, handmade journal aesthetic with pastel colors, washi tape, torn kraft labels, doodled arrows, hearts, flowers, clouds, stars, butterflies, stamps, a compass, suitcase, camera, tickets, map pin, and small panda doodles.

Main text hierarchy: At the top, large red-orange brush lettering reading {argument name="main Chinese headline" default="成都达人计划"}. Directly underneath, a beige torn-paper strip with handwritten English text reading {argument name="English subtitle" default="Chengdu City Guide"}. Below that, a longer torn-paper strip with Chinese text reading {argument name="Chinese guide title" default="成都旅游城市漫游计划"}, with small tag-like labels at the left. On the upper right, add a small speech bubble reading “City Tank Chengdu.” Add a beige note on the right reading “Best travel time: March–June | September–November,” styled like handwritten travel advice.

Layout and counted photo elements: Use exactly 16 main photo or cutout image elements arranged in a lively collage, each with tape or torn-paper label styling. The 16 elements are: 1) top-right Polaroid of a modern Chengdu skyline; 2) upper-left Polaroid of a cozy interior alley/teahouse labeled as Kuanzhai Alley slow life; 3) left-middle photo of a red lantern ancient street labeled Jinli Ancient Street; 4) center hero photo of three pandas in a tree labeled Giant Panda Breeding Research Base; 5) upper-right landscape photo of Dujiangyan river and mountains labeled Dujiangyan, World Heritage; 6) right-middle mountain photo labeled Qingcheng Mountain; 7) left-lower city street photo labeled Chunxi Road shopping/check-in spot; 8) left-lower panda cutout leaning from a building labeled IFS Panda landmark; 9) center photo of a gaiwan tea cup labeled People’s Park gaiwan tea and tea culture; 10) center-lower photo of a black steam train labeled Eastern Suburb Memory / East Park; 11) right-lower Polaroid of a night city skyline; 12) right-lower Polaroid of a warmly lit traditional street at night; 13) right-lower Polaroid of skyscrapers at night with a Sichuan opera face mask sticker; 14) bottom-left food photo of skewers labeled skewer aroma; 15) bottom-center large hotpot photo labeled Chengdu Food Map; 16) bottom-right food arrangement with tea, rabbit head, dumplings, and noodles.

Food section details: The bottom third should feel like a Chengdu food map. Include exactly 6 labeled food items: spicy red hotpot in the center, skewers on the left, dandan noodles at the bottom, rabbit dumplings on the right, rabbit head at the far right, and gaiwan tea above/right. Labels should be small torn-paper Chinese-style tags, but the overall prompt remains in English.

Visual style: Cute travel diary collage, semi-realistic printed photos mixed with illustrated stickers, soft grain, pastel blue background, cream paper scraps, coral-red title, muted orange labels, white photo borders, slightly rotated elements, visible tape strips in peach, blue, and green. The image should look like a handmade planner page or Xiaohongshu travel guide poster, dense but readable.

Constraints: Keep the poster vertical and information-rich. Maintain the exact count of 16 main image elements and 6 labeled food items. Do not add logos, watermarks, QR codes, or extra large text. Preserve a cheerful Chengdu tourism mood with pandas, tea, hotpot, historic streets, city skyline, and mountain scenery.

Create a cinematic vertical full-body street-fashion portrait of {argument name="character name" default="VLNS"}, a non-human alien character whose true biological head is a large glossy sphere, walking away in a modern city crosswalk at golden hour. The alien has no human face or skin: the head is a perfectly round, oversized helmet-like biological orb covered in vivid graffiti-camouflage patterns in turquoise, green, orange, pink, purple, and blue, with a jagged black monster-mouth graphic filled with purple triangular teeth on the side, circular ear-disc modules on both sides, and one large pink circular eye emblem with a black star pupil. The character wears a black punk bomber jacket with a thick gray-white faux-fur collar, chaotic neon paint splatter all over it, a large target emblem on the back, and a rectangular sleeve patch reading “VLNS.” Add black cargo jogger pants covered in neon paint splashes and alien/skull graphics, including a prominent green alien skull on the lower leg, plus hanging silver chains and small colorful alien-head charms at the hip. Include high-top glossy iridescent sneakers in purple, blue, and black with thick white soles and circular target logos. Pose the character mid-step from a low rear three-quarter angle, hands in pockets, confident and mysterious, occupying the center of the frame. Set the scene in a busy downtown financial district with tall reflective glass skyscrapers on both sides, blurred pedestrians in suits moving through the crosswalk, a green traffic light on the left, long shadows on wet-looking asphalt, and warm sun flare from the left background. Use photorealistic detail mixed with hyper-stylized fashion editorial energy, shallow depth of field, motion blur on background pedestrians, crisp focus on the character, high contrast, reflective surfaces, vibrant neon color accents against black clothing, no extra readable text except “VLNS,” no human head, no exposed human body parts.

Goal: Create a 3 by 3 comparison chart showing nine different AI image model outputs for the same character concept: a full-body Japanese straw scarecrow yokai warrior wearing a fox or oni-style mask, posed front-facing with arms stretched horizontally on a wooden crossbeam like a scarecrow.

Canvas: Vertical poster, 768 x 1152 aspect ratio, warm off-white parchment background, thin gutters between panels, each panel containing one full-body character illustration centered on a pale paper backdrop. Use a muted beige, straw, charcoal, navy, and red color palette.

Layout: Arrange exactly 9 rectangular panels in a 3-column by 3-row grid. Each panel has a dark label bar along the bottom edge with bold white sans-serif text; the first label bar is red to highlight the preferred model. Keep all labels readable and aligned left.

Text content: Use exactly these 9 panel labels, in this order from top-left to bottom-right: 1) {argument name="highlighted model label" default="Seedream 5.0 Pro *"}, 2) Seedream V4.5, 3) Krea v2 Large (NEW), 4) Flux 2 Max, 5) Hunyuan V3, 6) Kling O3 (4K), 7) {argument name="bottom left model label" default="GPT Image 2"}, 8) Nano Banana 2, 9) Grok Imagine.

Subject details: In every panel, depict the same core subject with variation in rendering style: a Japanese harvest-spirit scarecrow figure made of dry straw, ragged cloth, and wood, wearing a pale fox mask or demon mask with red markings, a straw cape or thatched mantle, dark indigo peasant trousers, rope bindings, sandals, a red cord accent, and a small brass bell hanging from the waist or chest. Include a curved sickle or hook-shaped blade at the side in several variations. The character should look like folklore concept art: eerie but charming, handmade, weathered, and ritualistic. The arms should be tied or resting along a horizontal wooden beam in most panels. Add subtle straw fragments, loose fibers, and small sketchy shadows near the feet.

Panel-by-panel variation: Top-left: delicate watercolor concept art, rounder straw cloak, darker face mask, small bell near the waist, misty gray wash behind the legs. Top-middle: cleaner anime-concept rendering with an orange fox mask, red scarf, blue-black outfit, straw shoulder cape, wooden post behind the head, sickle at the right hip. Top-right: simplified cute mask, broad straw poncho, tall straw hair plume, red bow cord at the neck, wooden beam visible across the shoulders. Middle-left: tall thin scarecrow with a pale mask, large circular halo-like straw frame behind the head, brass bell on the left side, long dangling red tassel. Middle-center: stockier figure with a white oni mask, layered straw skirt, both hands gripping the crossbeam, two curved sickles hanging from red cords. Middle-right: more dynamic rough painted version with sharp fox mask, straw hat-like hood, large bell on the waist, red small demon mask charm at the side, black balloon trousers. Bottom-left: grittiest and most textured version, dark muddy clothing, heavy ragged straw cloak, shadowy face mask, brass bell in front, scythe-like blade at the lower left. Bottom-middle: elegant refined figure with radiant circular straw halo, narrow fox mask, symmetrical straw cape, red neck cord, bell and hook at the waist. Bottom-right: friendly toy-like fox-masked scarecrow, compact body, neat straw cloak, wooden post behind head, indigo kimono-like clothing and straw shin guards.

Visual style: Hand-painted fantasy character-sheet comparison, Japanese folklore design, watercolor and gouache texture, ink sketch accents, high detail, warm parchment paper, slightly rough AI-model-comparison aesthetic. Make the illustration dense and textured, with visible fibers, dry brush marks, and subtle paper grain.

Constraints: Do not add extra panels, extra labels, logos, watermarks, captions, or background scenery. Keep every character full-body and centered. Ensure the exact count of 9 labeled model panels is clear.

A breathtaking high-fashion magazine cover featuring a young {argument name="ethnicity" default="Asian"} woman with delicate facial features and long flowing dark hair blowing in the ocean wind. She stands confidently on a rugged coastal rock, wearing an elegant {argument name="dress style" default="off-the-shoulder white lace wedding gown"} with intricate lace embroidery, sheer sleeves, and a flowing tulle and chiffon skirt that billows dramatically in the wind. Behind her stretches a vast {argument name="background" default="dark stormy ocean"} with powerful crashing waves beneath a sky filled with heavy, dramatic clouds. Numerous seagulls circle around her and soar across the sky, with several flying close to the camera in the foreground, creating depth through subtle motion blur while the woman remains perfectly sharp in focus. The composition is a medium-full shot in a vertical 3:4 aspect ratio, captured from a slightly low angle using the rule of thirds to emphasize her elegance, strength, and commanding presence. The scene is illuminated by soft diffuse overcast natural light with cool, moody low-key tones, creating a striking chiaroscuro contrast between the luminous white dress and the dark sea and sky. Every detail is highly realistic, including the rough texture of the coastal rocks, foamy ocean waves, intricate lace fabric, flowing dress movement, detailed bird feathers, and cinematic atmospheric clouds. The overall mood is melancholic, mysterious, powerful, dramatic, and evokes the calm before the storm. Styled as a luxurious modern VOGUE magazine cover with large elegant masthead typography at the top, ultra-realistic, editorial fashion photography, cinematic film look, premium luxury aesthetic, masterpiece, award-winning composition, ultra-detailed, 8K HDR.

Tight close-up levitation food shot, {argument name="food item" default="strawberry shortcake layers"} floating mid-air, extreme detail on whipped cream texture and fresh strawberry glistening surface, golden sponge crumb texture sharp, powdered sugar cloud frozen mid-burst, strawberry sauce drip frozen mid-fall, {argument name="background" default="pure black background"}, {argument name="lighting" default="dramatic single-source spotlight lighting"}, ultra high detail, vertical 4:5 --ar 4:5 --v 6

Extreme close-up levitation food shot, thick cream cake slice mid-air, golden honey drip thread hanging below catching dramatic light, blueberries and raspberries tumbling nearby, white frosting texture ultra sharp, powdered sugar cloud catching spotlight, pure black background, ultra high detail on honey thread and cream texture, vertical 4:5 --ar 4:5 --v 6

Tight close-up levitation food shot, golden pancakes floating mid-air showing detailed surface texture and toasted edges, butter cube mid-tumble catching warm light, fresh berries scattered at varying focus depths, powdered sugar cloud frozen in burst, maple syrup drip thread, rich warm brown background, dramatic single spotlight, ultra high detail on pancake texture, vertical 4:5 --ar 4:5 --v 6

A breathtaking monochrome watercolor portrait of a young woman, created with expressive black and gray ink washes on textured handmade watercolor paper. Ultra-realistic facial features with soft cinematic lighting, luminous skin, thick natural eyebrows, smoky eyeshadow, long eyelashes, full matte lips, and loose windswept dark hair with delicate flyaway strands. The portrait blends seamlessly into abstract watercolor splashes, ink bleeds, paint drips, and expressive brush textures around the edges. Fine {argument name="accent color" default="gold"} metallic cracks and subtle {argument name="style" default="kintsugi-inspired"} accents run naturally across parts of the face, neck, and background, glowing softly against the grayscale palette. High contrast between light skin tones and deep charcoal shadows creates a dramatic editorial atmosphere. The background is minimal white paper filled with organic watercolor blooms, ink splatters, paint drips, and artistic stains, giving the piece a raw fine-art appearance. Soft depth, painterly transitions, elegant composition, emotional expression, museum-quality illustration, luxury contemporary watercolor art, fashion editorial aesthetic, highly detailed, ultra sharp, 8K, masterpiece, award-winning digital watercolor painting.

Create an ultra-premium dreamy papercut storybook poster for “How to Train Your Dragon”, blending the tenderness of a healing parent-child picture book with the handcrafted wonder of a luxury pop-up book. The image must feel magical, warm, gentle, emotionally comforting, and beautifully dimensional, like a collector’s edition children’s book designed for intimate reading and lifelong keeping. The overall composition should be soft, elegant, and heartwarming, with the open book as the absolute visual anchor. Use a vertical composition with a beautifully opened premium storybook placed at the center. The book must feel luxurious and tactile: thick textured paper, high-quality hardcover structure, soft deckled page edges, subtle paper grain, elegant binding, and refined print finish. The open spread must bloom upward into a layered paper-sculpture world, as if the story is quietly unfolding from the pages. The core concept is: the world of “How to Train Your Dragon” emerging from a dreamy layered paper book, where a childlike Viking hero and a dragon meet in a moment of trust, curiosity, and gentle companionship. The tone must feel less like battle and more like friendship, wonder, and emotional safety. At the center of the pop-up scene, create a stylized young Viking child in a premium hand-drawn paper-doll aesthetic: large expressive eyes, soft facial features, small adventurous body proportions, layered Nordic clothing, tiny boots, gentle cape details, and a brave but tender emotional presence. The character should feel like a beautifully illustrated children’s book figure turned into a dimensional papercraft sculpture. Opposite the child, create a dragon rising from the book in a soft, magical pop-up form. The dragon should feel lovable, protective, and wondrous: elegant wing structure, rounded expressive face, deep midnight blue or soft blue-black body tone, painted paper texture, gentle motion arc, and warm emotional presence. The dragon must feel like a fusion of hand-drawn illustration, papercut sculpture, and storybook magic, not like a horror creature and not like a plastic toy. The emotional interaction is the heart of the poster: the child and dragon face each other in a quiet, tender first-bond moment, with a tiny glowing light, ember, or magical spark between them. This focal point should feel like trust being born. The relationship must feel healing, safe, and full of wonder, suitable for a premium parent-child picture book. Build the surrounding world entirely from layered paper engineering: stacked paper mountains, soft cliff contours, Nordic village rooftops, cloud cutouts, small waves, flying page layers, paper forests, and delicate sketch lines transitioning into dimensional structures. The world should feel handcrafted and immersive, with visible paper depth and elegant layer separation. Some areas remain flat illustration on the page, some rise as folded pop-up elements, and some become fully sculptural papercut scenery. The visual style must combine: luxury pop-up book design, soft papercut layering, dimensional hand-drawn illustration, and a warm healing picture-book atmosphere. Everything must feel artisanal, magical, and emotionally premium. Background must remain minimal and softly luminous: warm ivory, light parchment, pale cream, or soft warm beige, with gentle atmospheric blur and no clutter. No room setting, no shelves, no unnecessary props. The focus must remain entirely on the storybook and the world unfolding from it. Lighting must be extremely soft and refined: warm storybook glow, gentle golden-beige illumination, delicate shadows cast by layered paper edges, soft highlight on the child’s face, subtle luminous contour on the dragon wings, and a calm magical falloff across the book. The lighting must feel comforting, dreamy, and premium, never harsh or synthetic. Typography should be minimal, official, and beautifully integrated. Use: main English title: “HOW TO TRAIN YOUR DRAGON” small Chinese subtitle: “How to Train Your Dragon” Optional small supporting English line: “A DREAMY POP-UP TALE OF FRIENDSHIP” Typography should feel gentle, storybook-like, and premium, with elegant spacing and clean hierarchy. Include a minimal official-style logo treatment only. Avoid cluttered poster text, cheap family-film fonts, or excessive decorative effects. Color strategy: warm parchment, soft ivory, creamy beige, misty blue, deep dragon blue-black, muted forest green, dusty Nordic gray-blue, and a tiny warm ember glow at the emotional focal point. The palette must feel clean, soft, magical, and soothing. Overall mood: dreamy, healing, warm, dimensional, handcrafted, magical, tender, premium, storybook-like, unforgettable. Rendering style: ultra-detailed luxury pop-up picture book design, layered papercut story world, dimensional hand-drawn cartoon illustration, open collector’s edition storybook, stylized Viking child and dragon interaction, warm magical light, soft editorial lighting, minimalist premium background, healing fantasy picture-book poster, 8k, world-class visual quality. Negative prompt: flat illustration only, plastic toy look, cluttered background, messy collage, low-detail paper folds, generic dragon, horror monster, harsh cinematic darkness, overly realistic human, cheap children’s book design, loud typography, chaotic composition, neon colors, low-end cartoon rendering, random props, weak pop-up structure