GPT Image 2 prompts

Goal: Create a realistic mobile Discord screenshot showing an image-generation channel for an anime project called {argument name="project title" default="After Echos"}, with a dark theme interface and a vertical feed of storyboard scene images.

Canvas: Tall smartphone screenshot, portrait orientation, about 670×1200 pixels, dark charcoal Discord UI. Crop the bottom slightly so the fourth image is only partially visible.

Header: At the top, show the Discord channel header with a back arrow on the far left, a red notification badge with the number 1, a hashtag icon, the channel name {argument name="channel name" default="codex-images"} in bold white text, a small chevron, a green online dot, Japanese status text reading 「1人がオンライン」, and a circular search icon on the right.

Layout: Below the header, show a scrollable feed with exactly 4 rounded-corner image posts stacked vertically. Each post has a cinematic anime scene image followed by a gray filename-style caption. Use narrow side margins, rounded image corners, and Discord-like spacing.

Post 1: A dark, moody, wide cinematic anime image of a cluttered underground workspace or safehouse at night. One person sits at a desk facing multiple glowing computer monitors. The room has shelves, boxes, a couch, a low table, scattered objects, warm desk lamps, teal monitor light, and a gritty industrial atmosphere. Caption: “AfterEchos SB-S07 v004 / S07-C040”.

Post 2: A daylight anime street scene with exactly 2 main characters walking side by side: a tall young man with messy black hair, glasses, black shirt, and long navy coat; and a blonde young woman with hair tied back, wearing a pale hoodie, light pants, and a blue embroidered denim jacket. Urban pedestrians and storefronts recede in the background; the right side is shaded by a wall with greenery. Caption: “AfterEchos SB-S08 v004 / S08-C030”.

Post 3: A warm indoor anime conversation close-up with exactly 2 visible characters: the same black-haired young man in glasses and navy coat looking serious in the foreground, and an elderly white-haired man with glasses, beard, and a beige turtleneck seen from behind/right profile. Background shelves contain books or records under warm lamplight. Caption: “AfterEchos SB-S10 v004 / S10-C020”.

Post 4: Partially visible at the bottom, an anime vehicle interior with exactly 3 visible characters: the elderly white-haired man with glasses on the left, the blonde young woman in the center rear, and the black-haired young man with glasses on the right. City buildings are visible through the windows. A floating white circular down-arrow button overlaps the lower-right area.

Visual style: The embedded images should look like polished Japanese anime concept art with cinematic lighting and detailed backgrounds, while the surrounding interface should look like a real Discord mobile app screenshot. Keep all UI text crisp and legible, with no extra watermarks or unrelated icons.

Goal: Create a clean photographic character reference sheet for {argument name="character name" default="a young East Asian woman"}, designed like a model identity sheet for consistent AI character generation.

Canvas: Wide horizontal 16:9 layout on a pure white studio background, divided by thin light-gray grid lines into multiple rectangular panels. Use soft even studio lighting, realistic skin texture, neutral color grading, sharp editorial photography, and no decorative background elements.

Layout and count: Include exactly 12 visible character views/panels. Panel 1 is a very large close-up frontal portrait occupying the entire left third of the sheet. Panels 2 and 3 are upper head-and-shoulders three-quarter angled portraits in the top center. Panels 4 and 5 are left and right side-profile head portraits in the middle center. Panels 6, 7, and 8 are full-body standing views on the top right: front view, side view, and back view. Panels 9, 10, 11, and 12 form a bottom row of facial-expression portraits: neutral, slight smile, broad smile showing teeth, and surprised parted-lips expression.

Visible labels: Add small black Japanese section labels near the top-left of relevant grid areas, using {argument name="section labels" default="正面, 斜め, 横顔, 全身, 表情"}. Keep the labels minimal, plain sans-serif, and unobtrusive.

Subject details: The woman has long {argument name="hair color" default="black"} wavy hair parted near the center, soft natural makeup, fair skin, large dark eyes, delicate eyebrows, a small straight nose, and muted rose lips. Her expression is calm and elegant. Maintain the same face, hair length, facial proportions, and styling consistently across all 12 panels.

Wardrobe and accessories: She wears a fitted {argument name="top color" default="white"} long-sleeve scoop-neck knit top, high-waisted wide-leg dark charcoal trousers, black pointed shoes, a short pearl necklace, and small pearl earrings. In the full-body panels her posture is upright with arms relaxed at her sides.

Constraints: Make it look like a real studio photo reference sheet, not anime or illustration. Avoid extra people, props, logos, watermarks, heavy shadows, colorful backgrounds, or additional text. Keep the full-body back view showing long hair falling down the back and the trousers clearly visible.

Create a highly detailed cinematic anime illustration of a cyberpunk feudal Japan city. In the foreground, a young elf-eared anime girl with {argument name="hair color" default="short pale blue bob hair"} and red eyes sits calmly on the edge of a dark tiled temple roof, legs forward, one hand resting behind her. She wears an ornate black-and-crimson futuristic samurai kimono armor outfit with layered sleeves, straps, belts, small hanging talismans, red trim, floral embroidery, mechanical details, white thigh-high socks, and black armored geta-like footwear. Add a small red flower hair clip near her bangs. Beside and beneath her lies a long black-and-red mechanical polearm/naginata with a polished curved blade extending toward the lower left foreground. Behind her, show an enormous rust-red mechanical samurai bird or crane-like war machine towering over the city, made of layered armor plates, exposed pistons, cables, gears, rivets, glowing red optics, and a long blade-like beak or weapon arm; it should feel ancient, damaged, and industrial. The background is a vast dense dystopian Japanese megacity with pagoda rooftops, scaffolded towers, vertical signs with stylized kanji-like markings, misty skyscrapers, smokestacks, and a large multi-tiered cyber-temple on the right. Use a cool gray overcast sky, drifting debris and tiny birds, atmospheric haze, deep perspective, and muted desaturated colors with strong crimson accents. Composition: wide 4:3 landscape, character slightly right of center, giant mech on the left, city filling the horizon, roof tiles leading diagonally into the scene. Style: ultra-detailed anime concept art, intricate linework, realistic lighting, gritty sci-fi texture, cinematic scale, sharp focus on the character, no modern UI, no watermark, no readable English text.

Create a vertical anime-style illustration of {argument name="character name" default="Kaga Sumire"}, a petite young girl sitting on sunlit stone steps in an overgrown mountain garden. She has short bobbed {argument name="hair color" default="silvery white hair"} with soft bangs, teal-blue eyes, and a small purple flower hairpin on the side of her head. She wears a loose plain white T-shirt tucked into high-waisted charcoal-gray cropped pants with a black belt, white ankle socks, and white sneakers. Pose her seated casually on the steps with one knee raised, one hand resting near her cheek in a thoughtful cute gesture, and the other hand beside a woven basket planter filled with green shoots. The setting is an old terraced stone ruin or garden stairway covered with lush ivy, vines, grasses, and small white wildflowers; include exactly one wicker basket planter, one visible purple hair flower, and a broad wall of heart-shaped ivy leaves climbing up the right side. Behind her, show layered stone terraces with dangling vines, evergreen trees, misty blue mountains, and a pale warm sky with soft clouds. Use golden-hour sunlight from the upper left, strong rim lighting on leaves and hair, long soft shadows, warm highlights, cool blue-green distant haze, delicate linework, rich cel-shaded anime rendering, cinematic composition, peaceful nostalgic mood, highly detailed foliage and masonry, no text, no watermark.

Create a vertical black-and-white manga illustration of a serene, androgynous young Japanese musician playing a koto at night. The character has {argument name="hair color" default="silver-gray"} long layered hair with loose bangs partly covering the eyes and one thick braid draped over the right shoulder, delicate facial features, downcast closed eyes, and a quiet melancholy expression. They wear a traditional {argument name="outfit" default="light kimono with a dark obi and a translucent floral haori"}, with soft folds and subtle flower patterns on the sleeves. Show both hands in the foreground actively plucking the koto strings, with long elegant fingers and black finger picks or rings on multiple fingers; the koto fills the lower third of the image with visible wood grain, bridges, and taut parallel strings receding in perspective. Set the scene in an old Japanese room or veranda with wooden beams, shoji-style framing, shelves and hanging scrolls in shadow, opening to a moonlit garden. Outside, include exactly one full moon, clusters of hydrangea flowers, leafy shrubs, and one small stone lantern. Use dramatic moonlight from the left, deep shadows, high-detail crosshatching, fine ink linework, grayscale tones, cinematic contrast, and a slightly low close-up perspective focused on the musician’s hands and sorrowful face. No modern objects, no color, no text, no watermark.

Create a widescreen cinematic science-fiction concept art scene in a bleak monochrome palette. The setting is the colossal interior wreckage of a destroyed futuristic megastructure or crashed spacecraft, viewed from a low ground-level perspective under an enormous dark ceiling slab that dominates the top two-thirds of the frame. Show one lone female figure standing slightly right of center in the distance, small against the architecture, with long black hair, pale skin, and a flowing dark-and-white layered outfit or coat, giving a quiet post-apocalyptic heroine mood. Surround her with shattered metal plates, collapsed floor debris, hanging cables, jagged beams, and broken hull fragments; include three massive curved structural ribs receding into the bright misty opening behind her. Strong backlight floods in from the left and rear, creating a white foggy exterior glow and deep silhouettes in the foreground. Use dramatic scale contrast, sharp angular sci-fi industrial forms, subtle dust haze, high detail, matte-painting realism, cold gray and black tones, no text, no watermark. Aspect ratio 16:9. Optional customization: the lone character is {argument name="character name" default="an unnamed dark-haired woman"}, the environment is {argument name="ruined structure type" default="a crashed alien megastructure"}, the mood is {argument name="cinematic mood" default="bleak post-apocalyptic solitude"}, the color palette is {argument name="color palette" default="cold monochrome gray and black"}, and the lighting is {argument name="lighting style" default="harsh white backlight through fog"}.

Goal: Create an ultra-premium photorealistic Nigerian small chops restaurant advertisement poster for social media, emphasizing abundant food, bold typography, and a luxury black-and-gold brand look.

Canvas: Vertical 4:5 poster, high resolution, dark cinematic background with warm amber bokeh and subtle restaurant lighting. Use a black-to-brown gradient backdrop and a rustic wooden table surface at the bottom.

Main headline area: At the top center, add a small line-art chef hat icon in yellow above the title. Use the large headline text {argument name="headline text" default="Delicious SMALL CHOP Big Satisfaction!"}. Make “Delicious” and “Big Satisfaction!” white handwritten script, and make “SMALL CHOP” huge distressed condensed uppercase yellow lettering. Add a thin white underline flourish under the script.

Supporting copy: Under the headline, centered in small white text, write {argument name="subheadline text" default="Tasty bites, made fresh for every moment."}. On the upper right, add a circular yellow stamp badge with dashed border reading “FRESH TASTY & HYGIENIC”.

Left benefit list: On the left side, create a section titled “PERFECT FOR:” in yellow. Include exactly 4 checked items with yellow checkbox icons: 1) Parties & Events, 2) Meetings, 3) Hangouts, 4) Every Occasion.

Hero food arrangement: Center the poster around a large white round plate piled high with Nigerian small chops, photographed in glossy appetizing detail. Include exactly 6 visible food categories on and around the plate: 1) golden spring rolls stacked on the left, 2) puff-puff balls in the upper center, 3) triangular samosas on the right, 4) sauced peppered chicken pieces across the front, 5) small fried round bites at the front center, 6) a white ramekin of red dipping sauce on the right. Behind the plate on the upper right, place a brown kraft paper cup filled with skewered glazed meatballs, with multiple wooden skewers sticking upward. Add small parsley garnish accents around the plate.

Promotional badge: On the lower left overlapping the plate, add a burgundy circular badge with a stitched/dashed border. Text inside should read {argument name="badge text" default="Small Chops, Big Love"} with a yellow heart icon.

Thumbnail strip: Directly below the plate, add exactly 4 rounded rectangular food thumbnail cards with thin yellow borders: 1) crispy fried chicken bites, 2) golden puff-puff with one cut open showing filling, 3) triangular samosas, 4) mixed small chops platter.

Feature bar: Near the bottom, add a wide rounded yellow information bar divided into exactly 3 columns with black icons and text: 1) cloche icon with “FRESHLY MADE SERVED HOT”, 2) thumbs-up icon with “QUALITY INGREDIENTS GREAT TASTE”, 3) phone handset icon with “PLACE YOUR ORDER WE DELIVER!”.

Footer: Bottom left, add a yellow rounded square phone icon and bold call-to-action text “ORDER NOW” above the phone number {argument name="phone number" default="0812 345 6789"}. Bottom center-right, add the phrase {argument name="footer slogan" default="WE SERVE TASTE, YOU CRAVE!"} with “YOU CRAVE!” highlighted in yellow. Bottom right, include a small yellow delivery scooter icon.

Visual style: High-end commercial food photography, strong contrast, crisp details, glossy sauces, golden fried textures, warm highlights, premium Nigerian catering poster aesthetic, clean spacing, bold readable typography, no watermark, no extra logos, no misspelled text.

Create a realistic vertical fashion-lifestyle photograph of a young East Asian woman in a bright minimalist perfume boutique, standing beside a glossy white display counter and smelling a small clear rectangular perfume bottle held near her nose. She has long {argument name="hair color" default="black"} softly layered hair with curtain bangs, natural makeup, fair skin, and a calm introspective expression while looking slightly down and to the left in three-quarter profile. She wears a fitted ribbed short-sleeve square-neck crop top in {argument name="top color" default="pale lavender"} and a high-waisted {argument name="skirt color" default="white"} A-line mini skirt. The boutique background is sleek and luminous, with glass shelves, cool white LED strip lighting, and many neatly arranged perfume bottles and reed diffusers in soft focus; include one black bottle on a clear acrylic riser on the counter in the foreground. Use clean whites, silver accents, pale pastel tones, shallow depth of field, soft diffused commercial lighting, realistic skin texture, elegant editorial composition, vertical 9:16 framing, subject centered slightly left, upper body and skirt visible, no text, no logos, no watermark.

Goal: Create a realistic vertical smartphone screenshot of the ChatGPT mobile app showing an image-generation result, in a clean iOS interface.

Canvas: Tall 9:16 phone screenshot, white background, rounded iPhone safe-area edges, black iOS home indicator at the bottom. Status bar at the top shows 9:33 on the left and exactly 4 right-side status items: cellular signal bars, Wi-Fi icon, battery percentage reading 51, and a battery shape.

Layout: Top navigation row contains exactly 3 controls: a hamburger menu icon on the left, a small compose/edit icon near the upper right, and a three-dot more menu at the far right. Below it, place one light-gray rounded user message bubble aligned slightly right. Under the bubble, add a small gray section label reading “思考”. Below the label, show one large generated photo card with rounded corners, then a horizontal strip of exactly 8 small thumbnail images. At the bottom, show one rounded input bar with a plus icon, placeholder text, microphone icon, and a blue circular voice button.

Text content: The user message bubble contains Chinese text: “{argument name="prompt text" default="Masochist系男友视角斜侧手持手机拍摄风格，风格清新干净，一种随性松弛氛围感，随手记录。绝美，表情随机；ar9:16\n\n连续10张单独图"}”. The bottom input placeholder reads “询问 ChatGPT”.

Main photo subject: The large generated image is a casual handheld portrait of {argument name="character description" default="a young East Asian woman with long dark hair"} sitting indoors on a green upholstered chair, smiling warmly at the camera. She wears {argument name="outfit" default="a beige ribbed tank top and loose white pants"}. The pose feels relaxed and candid, shot from a slightly high, tilted boyfriend-perspective phone angle. Her hands rest casually near her lap, her expression is bright and natural, and the lighting is soft daytime window light.

Environment inside the photo: A cozy apartment living/dining room with dark wood floor, sheer white curtains, darker side curtains, a window with greenery outside, and a dining table with tall dark chairs in the background. Keep the photo realistic, slightly imperfect, and spontaneous rather than studio-polished.

Overlay controls: On the large photo card, include exactly 2 translucent circular overlay buttons near the bottom: one left button labeled “编辑” and one right button with an upload/share arrow icon. The thumbnail strip contains exactly 8 small rounded previews showing alternate similar portraits of the same woman; the final thumbnail is selected with a subtle light outline.

Visual style: Photorealistic iOS screenshot, clean white margins, soft gray interface elements, accurate mobile app spacing, natural Chinese UI typography, no extra watermarks, no additional panels, no extra generated images beyond the 1 main image and 8 thumbnails.

Create a full-body standing character design sheet of a slim young anime man named {argument name="character name" default="unnamed young man"} on a plain white background, drawn in detailed black-and-white manga ink style with fine crosshatching and grayscale shading. He has pale skin, a delicate serious expression, large sharp eyes, and glossy {argument name="hair color" default="black"} medium-length hair parted near the center with long bangs partly covering one eye. Pose him facing forward with a slight contrapposto, head subtly tilted down, one arm relaxed at his side and the other hand tucked into his sleeve or pocket area. Dress him in a modern Japanese formal outfit combining traditional and contemporary elements: a dark kimono-style haori jacket with wide sleeves over a white collared shirt, a wrapped obi belt tied at the waist, long vertically pleated hakama trousers in medium gray, and polished black lace-up dress boots. The image contains exactly one character and exactly five main visible outfit pieces: dark haori jacket, white collared shirt, obi belt, pleated hakama trousers, and black lace-up boots. Use elegant fashion-illustration proportions, long legs, narrow shoulders, crisp garment folds, strong black highlights in the hair and shoes, and no text, props, scenery, borders, or watermark.

Create a pure black-and-white binary walkable-area mask for a top-down video game map. Use a square 1:1 canvas, about 768×768 pixels, with a solid black background and a single connected irregular white region representing traversable ground. The white shape should resemble a fantasy dungeon/island silhouette: a broad central vertical open area, jagged organic edges, narrow corridors and protruding rooms branching left and right. Include exactly 6 prominent side features around the central white mass: 1 large rounded oval platform on the upper left connected by a thin neck, 1 smaller mushroom-like platform on the lower left connected by a winding corridor, 1 jagged crescent-shaped protrusion at the upper left edge, 1 irregular upper-right chamber with three narrow black vertical cutouts inside it, 1 large black enclosed blob-like obstacle intruding from the right-center edge with a curled hook tip, and 1 round lower-right room-like protrusion partially separated by black negative space. Add several small triangular notches and pixel-stepped jagged edges, but keep the image strictly two colors only: pure white for walkable space and pure black for non-walkable space. No text, no grayscale, no antialiasing, no shadows, no perspective, no texture; make the edges slightly pixelated and mask-like as if intended for SDF or collision-map conversion.

Create a highly realistic vertical smartphone street portrait of a {argument name="subject" default="young East Asian woman"} posing close to the camera at night in an urban shopping street. She has long dark brown to black hair with soft straight bangs, glossy strands, and subtle pink neon highlights from a sign above; one eye is playfully winking while the other looks directly into the lens, with a gentle closed-mouth smile. Her right hand is raised elegantly in front of her lips and chin, showing long pale manicured nails. She wears small hoop and cartilage earrings, a thin black choker necklace with a small charm, a gray spaghetti-strap top, and an oversized black off-shoulder jacket slipping down to reveal one shoulder; a thin chain strap crosses her left side. The background is softly blurred with a closed metal shop shutter, concrete storefront walls, a glowing red-and-white neon sign with unreadable cursive lettering, and two distant pedestrians on the left, creating natural street depth. Use {argument name="lighting mood" default="cool night lighting mixed with pink neon glow"}, shallow depth of field, realistic skin texture, natural makeup with winged eyeliner and rosy lips, candid social-media photography style, high detail, 9:16 vertical composition, close crop from head to upper torso, no watermark, no extra text, no distorted hands or facial features.

Create a highly detailed vertical anime illustration of {argument name="character name" default="Sol"}, a slim young man with messy tousled {argument name="hair color" default="black hair"} and bright golden eyes, startled with his mouth slightly open as he juggles an extravagant dim sum feast. He wears an ornate Chinese New Year outfit: deep crimson silk changshan-style top with black collar and cuffs, gold frog closures, intricate gold embroidery, dangling red tassels, gold chains, flower-knot ornaments, and matching dark red trousers. He holds two bamboo steamer baskets, one large basket in the foreground packed with assorted dim sum and one smaller basket to the right with buns and dumplings. Show exactly 13 airborne food pieces around him: 3 white soup dumplings, 2 pleated white buns, 2 shrimp dumplings, 2 round siu mai, 1 brown chocolate-like bun in the hamster’s paws, 1 golden fried bite, 1 crescent dumpling, and 1 small cylindrical custard tart. Add a giant round golden-brown hamster floating near his shoulder, cheeks puffed, tiny paws holding and eating the brown bun, with an innocent happy expression. The background is a warm festive Chinese restaurant or banquet hall with red paper lanterns, red lattice window patterns, hanging knot ornaments, plum blossom branches, soft golden sunlight, and a red diamond paper decoration with the Chinese character 福. Use dynamic composition, food flying in midair, rich red and gold palette, delicate line art, glossy anime rendering, high detail fabric texture, shallow depth of field, lively celebratory mood, no watermark, no extra readable text besides 福.

Create a high-end studio portrait of {argument name="subject" default="a confident adult man with medium tan skin"} from the chest up, centered against a seamless deep black background. He has a sharp short buzz cut with slightly longer spiked texture on top, a clean fade at the sides, thick dark eyebrows, a full well-groomed black beard, and a calm serious expression with closed lips, facing slightly to the viewer’s left. He wears glossy black Ray-Ban style sunglasses with black lenses, a black front frame, red temple arms, and a subtle city skyline reflected across both lenses. Dress him in an elegant all-black outfit: exactly 2 visible clothing pieces, a fitted ribbed black turtleneck sweater and a tailored black blazer with lapels. Use warm cinematic key lighting from the front-left to sculpt the face and beard, gentle rim highlights on the hair and shoulders, realistic skin texture, crisp facial hair detail, premium fashion editorial mood, shallow depth of field, ultra-realistic digital photography look, vertical 3:4 composition, no text, no watermark, no extra accessories.

Create a cinematic anime-style wide shot of {argument name="character name" default="a teenage boy"} riding a bicycle straight toward the viewer on a sunny riverside road in a modern Japanese city. The camera is positioned low and centered in front of the bike, close enough to show the handlebars, brake cables, gloved hands, front wheel top, and the rider’s upper body in a dynamic symmetrical composition. The rider has {argument name="hair color" default="deep red"} tousled hair, sharp anime eyes, a focused confident expression with a slight open-mouth smile, and wears a glossy {argument name="helmet color" default="red"} vented cycling helmet, black fingerless gloves, a black zip-up hoodie over dark clothing, navy knee-length shorts, and a black backpack with visible shoulder straps. The bicycle is a dark mountain bike or hybrid bike with flat handlebars and a blue-black frame. The setting is a bright river embankment road with a metal guardrail on the left, calm blue water beside it, a long steel truss bridge in the background to the left, apartment towers and city buildings to the right, green grass along the roadside, and a vivid blue sky with large fluffy white clouds. Use clean high-quality anime character rendering blended with a realistic detailed background, strong midday sunlight, crisp shadows cast onto the asphalt, vibrant colors, depth perspective, and a sense of forward motion. No text, no logos, no extra people, no vehicles, no watermark.

Create a luxurious vertical anime playing card illustration featuring {argument name="character name" default="an original pink-haired magician girl"} as the unique Queen of Spades. The canvas is a tall rounded-corner poker card with a cream background and an ornate gold filigree border, deep navy enamel panels, tiny star motifs, blue jewel accents, and pink rose decorations in the top-right and bottom-left corners. Put a large navy “Q” and spade symbol in the top-left corner and a matching rotated “Q” with spade symbol in the bottom-right corner. In the center, show a cheerful idol-magician anime girl in a dynamic seated pose reaching one hand toward the viewer while holding a decorated playing card in her other hand. She has {argument name="hair color" default="soft pastel pink"} twin-tail hair with braids, ribbons, small flower clips, bright blue eyes, rosy cheeks, and a gentle smile. Her outfit is an elaborate pink, white, and navy stage costume: tiny tilted top hat with ribbons and gold trim, frilly blouse, striped pink corset bodice, navy bow with gemstone brooch, gold chains, layered ruffled skirt, thigh-high white stockings, garter charm, lace cuffs, and many small heart and jewel accessories. The background inside the card frame is a magical casino-stage scene with glowing confetti, floating playing cards, triangular bunting, balloons, crystalline shards, sparkles, and soft pastel lights in pink, lavender, blue, and gold. Use extremely detailed shoujo anime rendering, glossy highlights, delicate line art, ornate fantasy card design, luminous atmosphere, high saturation but soft pastel palette, dramatic depth, and premium collectible card quality. Keep exactly one main character, exactly two visible corner index sets, and exactly two rose clusters; no extra text besides the card rank and suit.

Create an ultra-realistic macro photograph of a single artist paintbrush applying thick oil paint onto a stretched canvas. The brush enters diagonally from the upper right toward the center, with a {argument name="brush handle color" default="dark green"} wooden handle, a worn shiny silver metal ferrule, and natural bristles loaded with {argument name="paint color palette" default="deep navy blue and teal blue"} paint. Show exactly one paintbrush and one canvas surface. The canvas is angled in perspective, filling the frame, with visible beige raw fabric along the lower right edge and heavy impasto paint across the rest of the surface. Emphasize raised ridges, palette-knife-like texture, bristle grooves, wet glossy highlights, and small scraped edges in the paint. Use dramatic realistic lighting from the upper left, creating a dark soft shadow under the brush and crisp specular reflections on the metal ferrule and wet paint. Background should be dark and out of focus, with shallow depth of field, high-detail texture, natural color grading, and a cinematic close-up composition in a {argument name="aspect ratio" default="16:9 horizontal"} frame. No text, no hands, no extra tools, no watermark.

Goal: Create a tall single-page educational science infographic explaining {argument name="headline text" default="WHY THE SKY IS BLUE — AND SUNSETS ARE RED"} with the subtitle {argument name="subtitle text" default="Rayleigh scattering explained in one page"}. Use a clean flat-design science-magazine look, suitable for a classroom handout.

Canvas: Vertical 9:16 poster on an off-white background, framed by black horizontal bars at the very top and bottom. Center the white text “GPT Image 2” inside each black bar. Main infographic area has generous margins, thin navy divider lines, rounded cards, and a navy/teal/coral/yellow palette.

Layout: Use exactly 5 numbered sections, each marked by a dark navy circular badge with white numbers 1 through 5 on the left. Keep all typography geometric sans-serif, with bold uppercase navy headings and teal italic subtitle.

Section 1 — Visible Light Spectrum: Heading reads “VISIBLE LIGHT SPECTRUM”. Show one long rounded horizontal rainbow gradient bar from violet/blue at left through cyan, green, yellow, orange, and red at right. Add exactly 4 wavelength labels beneath it: “400 nm”, “500 nm”, “600 nm”, and “700 nm”, with small tick marks.

Section 2 — Sunlight Entering the Atmosphere: Heading reads “SUNLIGHT ENTERING THE ATMOSPHERE”. Draw a large stylized yellow sun on the left, Earth’s curved edge on the right, and a pale blue atmospheric band between them filled with many small blue air-molecule dots. Show exactly 4 blue horizontal light rays entering and scattering with small blue arrows, plus exactly 3 red horizontal rays passing straight through toward Earth. Include exactly 6 larger molecule/star scattering icons in the atmosphere. Add two explanatory captions below the rays: “Blue light: scattered ~5.5x more” in blue and “Red light: passes through” in red.

Section 3 — Formula and Bar Chart: Create two rounded rectangular cards side by side. Left card shows the large equation “I ∝ 1/λ⁴” and the caption “Scattering intensity is inversely proportional to the fourth power of wavelength”. Right card title reads “RELATIVE SCATTERING (450 nm vs 650 nm)”. Include a simple bar chart with exactly 2 bars: a tall blue bar labeled “450 nm” and a shorter red-orange bar labeled “650 nm”, y-axis labeled “Relative scattering”, and large text “≈ 4.4 : 1” on the right.

Section 4 — Noon vs Sunset: Make two large circular/oval diagrams side by side. Left diagram title: “NOON — short path through atmosphere”. It shows a small sun above a curved Earth horizon, a tiny black human silhouette at ground level, blue dots in the sky, and vertical red and blue rays passing through a short atmospheric path with blue scattering arrows. Right diagram title: “SUNSET — long path”. It shows a low sun near the horizon on the left, a long horizontal path through a pale blue atmosphere over curved Earth, blue light scattering away along the path, a red ray continuing to a black human silhouette at the right, and the caption “Blue is scattered away before reaching your eye”.

Section 5 — Facts: Create exactly 3 rounded fact cards in a row. Card 1 has a simple outlined Mars-like planet icon and text “Fact: Mars sunsets are blue”. Card 2 has a teal ocean wave icon and text “Fact: The ocean is blue for a different reason”. Card 3 has a purple eye icon with rays and text “Fact: Violet scatters most, but our eyes favor blue”.

Footer: Add small centered source text: “Sources: Rayleigh (1871) · Illustration for classroom use · Fig. 1 of 1”.

Visual style: Flat vector illustration, thin navy outlines, soft gradients only where needed, crisp readable labels, science textbook clarity, no photorealism except the Earth texture may be lightly detailed. Keep spacing balanced and ensure every label is legible.

Goal: Create a watercolor storyboard page showing a quiet, mysterious delivery story starring {argument name="character name" default="an elderly woman courier"} and her dog.

Canvas: Square image with a 3x3 storyboard grid on white paper, thin black gutters, soft watercolor-and-ink illustration. Add a black horizontal band at the top and bottom, each with centered white text reading “Qwen Image 3.0”.

Visual style: Gentle rainy-city watercolor, muted grays and greens, loose ink outlines, softly washed backgrounds, slightly vignetted panel edges, cinematic but hand-painted. Keep the same protagonist in every relevant panel: a woman in her 60s with short silver hair, round red glasses, a {argument name="raincoat color" default="mustard yellow"} raincoat, dark gloves, boots, and a small brown dachshund on a blue leash. The mood is quiet, suspenseful, and melancholic.

Layout: Exactly 9 panels in a 3 by 3 grid. Each panel has a small white circle with a black outline and black number at the top-left corner, numbered 1 through 9. Use thin black dividers between panels. Some panels include small caption boxes in a handwritten typewriter-like font.

Panel-by-panel content, exactly 9 panels:
1. Numbered 1. Wide view at dawn in heavy rain: the elderly woman stands at the green front door of an old townhouse, reaching toward the doorbell or knob, with her brown dachshund beside her on the wet pavement. Georgian-style street in the background, reflective puddles, iron railings. Caption box: “7:02 AM — The last delivery.”
2. Numbered 2. Close-up of her gloved hands holding a brown paper parcel tied with red string. The parcel label reads {argument name="parcel label" default="Unit 4B, Alder Lane"}. Rainy gray background.
3. Numbered 3. Street scene: the woman and dachshund wait near a curb in rain while a city bus passes in the background, headlights reflected in the wet road. Caption box: “The city never waits.”
4. Numbered 4. Low-angle view of the old apartment building facade: pale stone walls, tall windows, green door, iron fence, rain falling. The woman looks up at the building from the lower left, her red glasses visible.
5. Numbered 5. Close-up of a metal apartment buzzer panel with exactly six visible buttons labeled 1A, 4B, 2A, 3B, 4A, and one partially visible lower button. A muddy gloved finger presses the 4B button.
6. Numbered 6. Dark, narrow door opening: a young man’s face peers through the crack from inside, half hidden in shadow, anxious expression. Caption box: {argument name="whisper caption" default="You’re late,” he whispered."}
7. Numbered 7. Close-up of the parcel being handed forward against a dark rainy doorway, red string tied in a neat bow, water streaks in the background.
8. Numbered 8. Rear view of the elderly woman walking away down the rainy street with her dachshund on the blue leash. Empty city sidewalk, reflective pavement, quiet townhouses and trees fading into mist.
9. Numbered 9. Final static shot of the same green townhouse door from outside, now closed, with the red string hanging from or tied near the door. Wet pavement and sparse plants by the railings. Caption box: {argument name="final caption" default="But she never delivered anything at all."}

Constraints: Keep the protagonist consistent across all panels. Use exactly 9 numbered panels and exactly the captions specified. Do not add extra speech bubbles, extra characters, or extra panels. Maintain a mysterious rainy morning atmosphere and a hand-painted storyboard look.

Create a cinematic vertical portrait of {argument name="character name" default="a brooding Tokyo private detective"} standing on a wet rooftop at sunset, framed from mid-thigh upward in a 9:16 composition. He is a Japanese man in his early-to-mid 30s with medium-length messy black hair, light stubble, tired intense eyes, and a serious noir expression, looking directly at the camera with both hands in his trouser pockets. Outfit: exactly 5 main visible clothing/accessory pieces — a wrinkled pale gray button-up dress shirt with sleeves rolled to the forearms, a loosened dark tie, black shoulder holster straps, brown high-waisted pleated trousers, and a dark belt. The setting is a gritty urban Tokyo skyline after rain, with reflective puddles on the rooftop floor, a metal guardrail, dense city buildings with lit windows and neon signs below, and exactly 1 large rooftop billboard on the right reading 「東京探偵事務所」. Use dramatic golden-orange sunset light breaking through dark storm clouds, strong rim lighting around the hair and shoulders, deep shadows on the face, moody high-contrast color grading, realistic film photography, subtle rain-slick textures, 35mm cinematic lens feel, shallow depth of field, detailed skin and fabric, no extra characters, no modern sci-fi elements, no watermark.

Create a cinematic vertical portrait of a gritty Japanese noir detective on a rain-slick rooftop in Tokyo at sunset. The subject is {argument name="character description" default="a rugged middle-aged Japanese man with wet swept-back black hair, stubble, a stern tired expression, visible neck and chest tattoos"}, standing three-quarter frontal and looking slightly down at the viewer with a tough, brooding attitude. He wears {argument name="outfit" default="a dark pinstripe suit, black open-collar shirt, leather-like dark coat, belt, and loose trousers"}; one hand is in his pocket and the other holds a cigarette near his thigh. The environment is a dense urban Tokyo skyline with glowing windows, neon signs, and industrial rooftop details, all wet from rain with strong reflections on the floor. Include exactly 3 prominent signs: a large weathered rooftop billboard on the right reading 「東京探偵 事務所」, a vertical red neon hotel sign on the left reading 「ホテル ニュー東京」, and a smaller red sign in the mid-background reading 「ポンコ」. Use {argument name="time and sky" default="dramatic golden-orange sunset breaking through dark storm clouds"} as the main backlight, creating rim light around the character and high-contrast shadows on his face and suit. Style should be ultra-realistic cinematic photography, moody neo-noir crime drama, 35mm film look, shallow depth of field, wet textures, volumetric haze, rich amber highlights and deep black shadows, high detail skin texture, realistic fabric folds, no cartoon styling, no extra characters, no watermark.

Goal: Create an ultra-detailed 16:9 fantasy character design sheet for {argument name="character name" default="LUNARIA"}, titled {argument name="character title" default="MOONLIGHT WITCH"}, in the style of an official AAA fantasy RPG character bible.

Canvas: Wide horizontal character sheet on an elegant ivory parchment background with a thin delicate gold decorative border, ornamental corner flourishes, soft pastel lighting, refined editorial spacing, and a luxury magical-girl fantasy aesthetic.

Main character: A petite anime moon witch girl with a gentle cheerful expression, pale skin, large rabbit ears, very long {argument name="hair color" default="white-to-blush pink"} hair in thick braided twin tails, warm pink eyes, and small ornate hair accessories including dark crescent-moon decorations, ribbons, and gold accents. Her outfit is a short white layered dress with translucent lavender underskirt, a large red bow at the chest, gold filigree trim, floral embroidery, pink tassels, and an oversized dark navy-black cape with huge kimono-like sleeves lined in rose pink. Include moon, rabbit, blossom, ribbon, bell, tassel, and soft magical sparkle motifs. She wears high platform sandals with rabbit decorations, ankle frills, and pink-lavender socks. Pose the main full-body illustration on the left, floating lightly with arms open and one leg lifted, as if presenting herself.

Layout: Top-left typography reads exactly “LUNARIA” in large elegant serif letters, with “MOONLIGHT WITCH” below in smaller serif text and a tiny decorative butterfly icon. Arrange the sheet into clearly labeled sections on the right and bottom while keeping the design airy and symmetrical.

Sections and exact counts: Include 1 large main full-body character illustration on the left. Include 3 smaller turnaround full-body views across the upper right labeled “TURNAROUND”: front view, side view, and back view. Include 3 accessory items in a rounded display panel labeled “ACCESSORIES”: one glowing rabbit moon lantern with tassel, one pink ribbon bow with twin tassels, and one gold spherical charm with tassel. Include 4 circular portrait expressions labeled “EXPRESSION”: happy open-mouth smile, neutral or shy pout, excited smile, and closed-eye content smile. Include 4 rectangular close-up panels labeled “DETAILS”: chest bow and bodice, layered skirt and waist ornament, embroidered cape sleeve with rabbit/floral motifs, and rabbit-ear hair accessory close-up. Include 6 square color swatches labeled “COLOR PALETTE”: coral pink, mauve purple, deep plum/navy, warm gold, cream ivory, and pale blush white. Include 4 rabbit companion drawings: one larger fluffy white rabbit sitting near the main character, plus a bottom rounded panel showing the rabbit in exactly 3 views: front, side, and back.

Style: Highly polished anime concept art, delicate linework, soft cel shading blended with painterly highlights, pastel pink and lavender accents, ornate gold detailing, crisp character-design-sheet presentation, clean readable labels, no clutter.

Constraints: Keep all text in English, preserve the visible labels “TURNAROUND,” “ACCESSORIES,” “EXPRESSION,” “DETAILS,” and “COLOR PALETTE,” use exactly the counted elements above, avoid watermark or logos, and keep the background mostly empty ivory so the character sheet remains readable.

Ultra-realistic cinematic portrait of a {argument name="subject" default="stylish young man"} sitting in the driver's seat of a sleek black luxury sedan with the window rolled down, wearing a {argument name="outfit" default="perfectly tailored black suit over a black turtleneck"}, black aviator sunglasses, and a premium stainless-steel wristwatch. He rests one hand on his chin in a confident, thoughtful pose, with a sharp fade haircut, voluminous textured hair, and a well-groomed beard. Luxury urban street in the background with modern buildings and warm bokeh lights, soft overcast daylight, shallow depth of field, natural skin texture, dramatic composition, premium lifestyle photography, 85mm lens, f/1.8, photorealistic, ultra-detailed, HDR, cinematic color grading, luxury fashion editorial, crisp focus, high contrast, 8K, masterpiece.

A {argument name="subject" default="luminous jellyfish"} suspended against a {argument name="background" default="pure black background"}, presented like a futuristic scientific specimen plate. Its translucent bell glows in {argument name="colors" default="saturated neon magenta, ultraviolet purple, electric cyan, cobalt blue, and toxic green"}, with branching vein-like structures and fine cellular textures visible inside. Long ribbon-like tentacles trail downward in elegant curves, each edge illuminated by fluorescent light. Along the left side, six clean rectangular swatches show gradients and macro texture studies derived from the jellyfish’s membranes. High-contrast ultraviolet photography, iridescent biological detail, crisp symmetry, minimal composition, deep black negative space, surreal natural-history aesthetic, intensely radiant neon color.