GPT Image 2 prompts

Reference images 1 and 2. [Person's Pose and Expression] The person is staring straight into the camera with a haughty and alluring gaze. Lightly and calmly wrapping both hands around a {argument name="bouquet color" default="huge crimson rose bouquet"} held as if holding a baby, creating a calm yet mysterious and faint atmosphere. [Person's Outfit] Wearing a {argument name="outfit style" default="black sleeveless dress"} with a luxurious and matte texture. The lower skirt part has voluminous and soft pleats, making the color of the gorgeous red roses visually stand out more. [Person's Hairstyle and Makeup Details] A brown-toned long wavy hairstyle with natural volume. Thick wavy bangs naturally flow along the face line, slightly covering one eye. The makeup highlights the lips by applying a vivid red lip matte on clean, white skin expression. [Lighting and Light Direction] Powerful natural light seeping from the left of the person softly wraps around the hair's gloss and shoulder line, shining powerfully enough to make the skin look whiter. [Texture and Color Mood] The deep black of the outfit, the intense red of the roses, and the velvet-like smooth texture of the rose petals contrast vividly with the clean skin texture, creating a seductive and classic fashion pictorial mood. [Film and Camera Lens Depth, Angle] A vertical view at eye-level from the front for a fashion magazine cover. A shallow depth of field shoot that focuses on the upper body of the person and the rose bouquet and softly out-focuses the background to make the subject stand out dramatically. A 9:16 ratio cowboy shot. [Background Elements (Optional)] An exotic architectural background with classic European-style outdoor stone stairs and flower pots containing green plants. A huge '{argument name="typography" default="BLANC"}' typography with a yellow and orange gradient is placed at the top of the screen. [Sign] The Threads logo operated by Meta is located small in the bottom right, and "CHAse" is written small above it like a signature in white cursive.

Transform the provided reference image into an ultra-premium futuristic point-cloud portrait. Preserve the exact pose, body proportions, clothing silhouette, facial structure, hairstyle, and overall composition from the reference image. Reconstruct the entire subject using millions of {argument name="particle color" default="glowing white"} volumetric particles, data points, voxels, and digital dust, creating a high-density holographic scan effect. Subject standing centered in a black void, surrounded by subtle floating particles and sparse {argument name="accent color" default="red"} digital noise. Generate a realistic 3D point-cloud body structure with particle trails, depth mapping, scan-line artifacts, and volumetric light dispersion.
Monochrome black-and-white color palette, extreme contrast, cinematic darkness, cyberpunk archive aesthetic, futuristic AI scanning system, digital twin visualization, neural network reconstruction, holographic particle rendering, volumetric point-cloud simulation, generative systems artwork, ultra-detailed particle density, glowing edge contours, subtle light bloom, depth fog, data stream particles falling vertically around the subject.
Add {argument name="ui metadata style" default="minimal technical UI typography"} around the frame such as archive labels, scan identifiers, model codes, generative system metadata, voxel render information, and futuristic research annotations. Use clean micro-typography with a luxury tech-brand aesthetic. Keep text minimal, small, and strategically placed near the edges.
Pure black background, center composition, museum-grade digital art, premium poster design, experimental AI visualization, ultra-realistic particle simulation, high-end CGI render, Octane render quality, Unreal Engine 5, volumetric lighting, ray tracing, sharp particle definition, cinematic glow, futuristic archive documentation style.
Style Keywords: Point Cloud Human, Digital Ghost, Neural Reconstruction, Voxel Scan, Data Visualization, Cyber Archive, Holographic Entity, Generative Systems, Monochrome Tech Art, Luxury Futurism, Volumetric Particles, AI Scan Portrait, Experimental CGI.
colorful background, cartoon, anime, low resolution, blurry subject, extra limbs, distorted anatomy, duplicate body parts, oversaturated colors, messy composition, excessive UI elements, watermark, logo, text overlap, low particle density, flat lighting, poor contrast, unrealistic proportions.

Create a high-end 4:3 {argument name="editorial style" default="luxury fashion editorial"} storyboard in 3x4 grid (12 frames), {argument name="fashion house style" default="Valentino/Bottega Veneta"} style, {argument name="color palette" default="ivory cream + deep burgundy"} palette. Structured flow: dressing room, street walk, rooftop pose, golden hour close-up, finale

Create a realistic premium bicycle helmet, cycling suit, and bicycle based entirely on the uploaded illustration. The helmet shape, visor design, aerodynamic structure, cycling suit silhouette, bicycle frame, wheels, materials, symbols, decals, lighting, textures, gloves, shoes, and mechanical details must all originate from the illustration’s atmosphere, colors, fashion, emotions, worldbuilding, fantasy or sci-fi elements, and character design. Do not create a generic sports bicycle or cycling outfit. Adapt the entire design language to the uploaded artwork itself. Shrine maiden themes may inspire elegant lacquer-like frame parts, sakura patterns, spiritual crests, layered fabric details, glowing talisman ornaments, and traditional Japanese aesthetics. Cyberpunk or android themes may inspire angular aerodynamic helmets, illuminated circuitry, holographic materials, segmented armor-like suits, neon accents, glowing wheel rims, and futuristic carbon textures. Cute pastel themes may inspire glossy rounded forms, transparent materials, soft gradients, playful accessories, and stylish fashion aesthetics. The helmet and bicycle may use exaggerated or surreal shapes naturally inspired by the illustration itself, such as mechanical cat ears, wing-like fairings, floating ornaments, asymmetrical structures, jellyfish-like transparent parts, or fantasy-inspired wheel designs. The bicycle itself must feel born from the same universe, with completely original frame geometry, wheel shapes, drivetrain styling, lighting systems, and fantasy technology. If the illustration contains a character, place the same character naturally beside or riding the bicycle while preserving the original face, hairstyle, proportions, colors, and personality. Bright cinematic lighting, premium futuristic bicycle advertisement photography, ultra detailed, realistic reflections.

Goal: Create a production-grade environment reference sheet for a {argument name="historical setting" default="Royal Gupta Empire palace complex, 5th century CE"}, designed as a cinematic storyboard and worldbuilding asset for GPT-image generation.

Canvas: Wide 16:9 reference-board layout on a dark charcoal background, dense but clean, with thin white divider lines, uppercase white technical labels, small annotation callouts, and a realistic architectural concept-art style. The overall title across the top reads: “PRODUCTION GRADE ENVIRONMENT REFERENCE - | ROYAL GUPTA EMPIRE GUPTA EMPIRE PALACE COMPLEX”, with “5TH CENTURY CE” aligned at the top right.

Main composition: Place a large central aerial three-quarter view of a grand Gupta-era palace complex occupying the middle of the sheet. The complex is made of red sandstone and cream stone, with symmetrical courtyards, high enclosing walls, arched gateways, colonnaded halls, carved balconies, flat terraces, small domed pavilions, lotus pools, geometric garden paths, fountains, mango trees, flowering shrubs, and warm golden-hour sunlight. Add fine white callout lines labeling architectural and landscape details such as carved columns, throne platform, mango gardens, marble seating, and sacred lotus ponds.

Left column sections: Include exactly 3 sections. Section 1 is “PALACE ENTRANCE”, showing a frontal red sandstone gate with two tower-like chhatri pavilions, a central pointed arch, two stone guardian statues, red-and-gold hanging banners, and relief carvings beneath it. Label exactly 3 entrance detail thumbnails below it: “Stone guardians”, “Banners”, and “Relief carvings”. Section 2 is “ROYAL GARDENS”, showing a formal garden with rectangular water channels, fountains, clipped lawns, trees, peacocks, and palace colonnades in the background. Section 3 is “MATERIAL REFERENCE”, showing exactly 8 material swatches labeled: Red sandstone, Cream sandstone, Polished granite, Marble inlays, Carved teak wood, Bronze fittings, Copper elements, and Painted plaster mural pigments and gold leaf.

Top and right sections: Include exactly 5 smaller architectural reference panels around the central image. Panel 1: “ROYAL AUDIENCE HALL”, an interior hall with carved stone columns, warm shadows, elevated audience platform, carved columns, and throne platform labels. Panel 2: “QUEEN’S COURTYARD”, a garden courtyard with colonnaded palace walls, flower beds, central fountain, marble paths, and lotus ponds. Panel 3: “MATERIAL REFERENCE”, another grid of exactly 8 swatches with the same labels: Red sandstone, Cream sandstone, Polished granite, Marble inlays, Carved teak wood, Bronze fittings, Copper elements, and Painted plaster mural pigments and gold leaf. Panel 4: “PALACE BALCONY SYSTEM”, showing carved projecting balconies with stone railings and ornate supports. Panel 5: “SACRED TEMPLE ZONE”, showing small Gupta-style shrines with domes, red stone, trees, and a courtyard.

Bottom row sections: Include exactly 3 sections. Section 1: “LIGHTING REFERENCE”, with exactly 4 cinematic lighting thumbnails labeled Dawn, Midday, Sunset, and Moonlit Night. Dawn should be pale and misty, midday bright and clear, sunset orange and dramatic, moonlit night blue with palace silhouettes. Section 2: “COLOR PALETTE”, with exactly 8 color chips labeled Sandstone Red, Warm Ivory, Lotus Pink, Deep Saffron, Bronze Brown, Peacock Blue, Banyan Green, and Royal Crimson. Section 3: “CULTURAL DETAIL”, with exactly 9 illustrated cultural elements labeled Royal elephants, Attendants, Sacred cows, Oil lamps, Incense burners, Flower garlands, Palm-leaf manuscripts, Bronze vessels, and Silk banners.

Atmosphere note: Add a small paragraph near the sacred temple panel describing warm Indian climate lighting, visible sunbeams in shaded courtyards, lotus pools, shisham/rosewood and neem trees, monsoon-season atmosphere, incense smoke, buffalo or sacred animals, rich bronze and gold ornamentation, birds, and subtle reflections.

Visual style: Ultra-detailed realistic environment design sheet, authentic ancient Indian Gupta architecture, red sandstone carvings, cream sandstone highlights, polished stone, carved teak, bronze and copper details, painted plaster with gold leaf, volumetric god rays, warm golden-hour ambience, crisp labels, cinematic consistency across camera angles.

Constraints: Keep the sheet information-dense but readable, with exactly the counted panels, material swatches, lighting thumbnails, color chips, cultural detail objects, and entrance detail thumbnails listed above. Avoid modern buildings, modern people, plastic materials, neon colors, sci-fi elements, watermarking, or random extra labels. Use the custom palette emphasis of {argument name="dominant material color" default="red sandstone"}, {argument name="secondary material color" default="warm ivory cream sandstone"}, {argument name="lighting mood" default="golden-hour volumetric sunlight"}, and {argument name="main environment feature" default="symmetrical palace courtyards with lotus pools"}.

Goal: Create a vibrant 2x2 poster grid of four miniature 3D city street-food stands, each shown as an isometric diorama on a small rounded sidewalk base, with bold city-and-food titles above each scene. The image should feel like collectible toy miniatures rendered in polished clay/plastic 3D, highly detailed, cheerful, and social-media friendly.

Canvas: Square 1:1 composition divided into exactly 4 equal quadrants by thick black cross-shaped dividers. Each quadrant has a solid bright background color and one centered food-stand diorama with a large playful title at the top.

Layout and sections: Use exactly 4 food stands:
1. Top-left: {argument name="Tokyo stand title" default="Tokyo Ramen Stand"}. Red background. A compact Japanese ramen shop with red lanterns, Japanese signage, a glowing red ramen sign, wooden storefront, noren curtains, stools with customers eating at the counter, small crates, a bicycle, cherry blossom accents, and a tiny ramen bowl icon under the title.
2. Top-right: {argument name="Mexico stand title" default="Mexico Taco Stand"}. Green background. A Mexican taco stall with terracotta roof tiles, a large sign reading “TACOS,” green-and-white striped awning, customers on stools, taco ingredients at the counter, cactus plants, a red pickup truck parked beside it, streetlamp, menu boards, and a small taco icon under the title.
3. Bottom-left: {argument name="NYC stand title" default="NYC Hot Dog Cart"}. Yellow background. A New York hot dog cart with metal cart body, blue-and-yellow umbrella, illuminated “HOT DOG” sign, condiment bottles, pretzels or buns, one customer standing in front, subway entrance railing, trash can, street pole with signs, pedestrian signal, and a small hot dog icon under the title.
4. Bottom-right: {argument name="Paris stand title" default="Paris Crêpe Kiosk"}. Sky-blue background. A Paris crêpe kiosk with dark green vintage booth, striped awning, sign reading “PARIS CRÊPES,” customers gathered around the counter, chalkboard menus, ornate streetlamp, flowers, a leafy tree, bicycle, wicker basket, bollards, and a small crêpe icon under the title.

Visual style: Cute isometric 3D render, miniature diorama / toy model aesthetic, high detail, soft ambient occlusion, warm tiny lights, crisp shadows, saturated colors, rounded edges, realistic but stylized materials, thick playful white or black display lettering depending on contrast, slightly puffy retro typography.

Text content: Titles must be large, legible, and centered near the top of each quadrant. Keep the main shop signs readable where possible: “TACOS,” “HOT DOG,” and “PARIS CRÊPES.” Japanese shop signs can use authentic-looking Japanese characters. Avoid adding extra captions outside the four titles and shop signage.

Constraints: Exactly 4 quadrants and exactly 4 diorama food stands. Keep each stand fully visible within its quadrant, with no overlap across the black dividers. No watermark, no logo, no photorealistic humans, no extra panels.

Goal: Create a cinematic 2×2 portrait grid titled by mood only, showing four anonymous legendary figures in dramatic red lighting, with their faces intentionally hidden by opaque rectangular censor blocks so they are recognizable only by silhouette, hair, posture, and styling.

Canvas: Vertical 3:4 image, four equal panels separated by thick clean white gutters forming a cross. Each panel is a tight chest-up studio portrait against a dark crimson-to-black background with a vivid red rim light or glowing red spotlight behind the subject.

Layout: Exactly 4 portraits total: top-left, top-right, bottom-left, bottom-right. Keep all subjects centered, cropped from upper chest to just above the head, wearing dark black coats or jackets over black shirts. Use high-contrast shadows, moody editorial photography, shallow depth of field, realistic skin and hair texture, and a premium magazine-cover look.

Subject details: Top-left portrait: an elderly figure with wild, voluminous white curly hair spreading outward, black overcoat, red halo light behind the hair. Top-right portrait: an older figure with short gray hair, dark beard or facial hair visible around the lower face, black coat, strong red side light. Bottom-left portrait: a bald middle-aged figure with glasses partly visible at the edges of the censor block, black coat, red rim light on shoulders. Bottom-right portrait: a younger-to-middle-aged figure with shoulder-length dark wavy hair, black coat, red light glowing behind the head.

Face concealment: Place exactly 4 opaque rectangular blocks, one centered over each face. The blocks should be vertical rectangles in warm brown skin-tone colors, covering the eyes, nose, and mouth while leaving hair, ears, neck, and clothing partly visible. Make the censor blocks flat, untextured, and deliberately artificial.

Style: Photorealistic studio portrait photography, dramatic red and black color palette, cinematic contrast, subtle film grain, sharp clothing edges, glowing red backlight, serious and iconic mood, no typography, no logos, no watermark.

Customizable direction: Use {argument name="color theme" default="deep crimson red and black"}, {argument name="portrait count" default="4"}, {argument name="subject theme" default="anonymous legendary figures"}, {argument name="censor block color" default="warm brown skin-tone"}, and {argument name="lighting style" default="dramatic red rim lighting"}.

Create a meticulously detailed vertical composite studio photograph showing the life progression of the same {argument name="gender" default="female"} person from birth to mature adulthood, with all ages sharing the same recognizable identity, facial structure, eyes, nose, lips, skin tone, and family resemblance. Canvas: portrait orientation, about 3:4, cinematic fine-art photography. Composition: arrange exactly 6 versions of the same person in a diagonal ascending timeline from the lower left foreground to the upper right background, each progressively larger and older, partially overlapping like a layered generational portrait. The 6 labeled stages must be: 1) “Newborn Baby (0 Months)” at the bottom left, swaddled in a soft white blanket and lying curled in a shallow nest-like wrap; 2) “Toddler (1–2 Years)” seated just above, wearing a navy-and-white striped long-sleeve shirt, curly dark hair, small bare legs visible; 3) “Cheerful School Kid (6–8 Years)” standing behind the toddler, wearing a navy polo shirt and carrying a pink backpack strap over one shoulder; 4) “Teenager (14–17 Years)” behind the school kid, with long dark hair, wearing a dark navy hoodie; 5) “College-Age Young Adult (20–30 Years)” behind and to the right, wearing a crisp white button-up blouse, confident neutral pose; 6) “Confident 57-Year-Old Adult (Adult)” tallest at the upper right, with long wavy salt-and-pepper hair, wearing a black blazer over a white collared shirt. Add small elegant white serif labels beside each figure on the left or near the corresponding head, matching the exact stage text above. Lighting and mood: dramatic Rembrandt-style studio light from the upper right, warm golden rim light on the right side, cool blue shadow tones on the left, soft haze and brown smoky atmosphere rising from the lower right. Background: painterly textured gradient backdrop transitioning from deep blue on the left to warm amber-brown on the right, with subtle vignette. Faces should be visible, realistic, and consistent across ages, with natural skin texture and age-appropriate hair, proportions, and clothing. Use photorealistic detail, shallow depth of field, premium portrait retouching, no extra people, no watermark, no border, no additional text beyond the 6 labels.

Create a high-fashion cyber-futuristic editorial portrait of a {argument name="subject" default="young woman with long dark hair"} in an icy chrome environment. The model is framed from the upper torso up, slightly leaning back and turned three-quarters toward camera, with the face deliberately obscured by a centered opaque square censor block in a muted gray-lavender tone. She wears a glossy silver translucent PVC corset or bustier with clear straps and reflective panels, plus one visible futuristic white-silver sleeve or glove on the raised right arm, decorated with small metallic studs, buttons, and dangling transparent tubes. Her right hand is lifted near her face with elegant long manicured nails and silver rings, while her dark hair falls over one shoulder. Surround her with exactly 6 prominent abstract chrome structures: 1 large blurred curved chrome blade in the left foreground, 1 thin arcing wire near the upper left, 1 horn-like chrome spike above the head, 1 intricate spiked chrome lattice across the upper right, 1 circular ring-like chrome loop at the right edge, and 1 soft translucent curved tube crossing the lower center foreground. Use a pale white and cool blue studio background, high-key lighting, soft bloom, shallow depth of field, glossy reflections, ethereal sci-fi fashion styling, cold metallic color grading, and a polished magazine-cover look. Keep the image text-free, watermark-free, and do not add extra people or props beyond the chrome abstract set pieces.

Goal: Create a realistic editorial magazine cover for an urban commuter lifestyle publication titled {argument name="magazine title" default="URBANIST"}, shot inside a crowded subway train during rush hour.

Canvas: Vertical magazine cover, 3:4 aspect ratio, full-bleed documentary street-photography background with overlaid magazine typography.

Main image: A candid full-body photo of a stylish office commuter standing in the subway aisle near the train doors. She wears a crisp white button-up blouse, a dark fitted mini skirt, sheer black thigh-high stockings, and glossy black pointed heels. She holds an overhead subway hand strap with one raised arm and carries a small black handbag in the other hand. Her face is intentionally pixelated/blurred for anonymity. Around her are seated and standing commuters in business attire, also with some blurred faces. The subway interior has metal poles, hanging hand straps, windowed sliding doors, route maps above the doors, warning stickers, fluorescent lighting, and a muted gray-green floor. The overall feeling is candid, metropolitan, slightly gritty, and fashion-editorial.

Layout: Large masthead across the top in oversized pink high-contrast serif capitals reading {argument name="magazine title" default="URBANIST"}. Above it, a small italic tagline reads “for commuters, by commuters.” On the upper right, small white issue text reads {argument name="issue date" default="APR. 2025"} on one line and {argument name="volume number" default="VOL. 47"} beneath it. Place cover lines along the left side and one on the lower right, using a mix of white serif headlines and smaller pink subtext. Add a barcode block at the bottom left.

Cover lines: Use exactly 4 editorial cover-line groups: 1) “MORNING RITUALS” with subtext “real routines, real people”; 2) “CITY FLOW” with subtext “stories from the underground”; 3) “OFFICE CODE” with subtext “weekday looks that work everywhere”; 4) “NEXT STOP” with subtext “where will your commute take you?”.

Barcode details: Bottom-left white barcode label with black bars, small text “URBANIST Vol.47,” issue marker “04>,” and digits resembling “9 771234 567003.”

Visual style: Photorealistic smartphone candid photo blended with polished magazine-cover design, natural subway lighting, slight motion blur outside the train window, muted cool gray environment contrasted with the soft pink masthead. Typography should feel like a modern fashion/lifestyle magazine: elegant serif display type, clean alignment, strong hierarchy.

Constraints: Keep exactly 4 cover-line groups, 1 barcode block, 1 main standing commuter subject, and 1 large masthead. Do not add extra logos, watermarks, QR codes, or unrelated text. Maintain anonymous blurred faces for visible commuters.

Using REFERENCE_0 as the source, extract and recreate only the transit poster artwork as a clean, front-facing digital poster, removing the subway wall, frame, glass glare, perspective distortion, wrinkles, reflections, and surrounding train interior. Preserve the original poster design and Chinese copy, but make it look like a professionally typeset vertical public-service poster.

Canvas: vertical 9:16 poster, full bleed, straight-on view.

Image treatment: keep the same central tourism photo concept from the reference: two hands forming a heart shape around the sun, with a soft pavilion and green lotus scenery in the distance. Make the photo cleaner, brighter, more cinematic, and centered, while retaining the same overall crop and mood.

Layout: top image section fills most of the poster; lower white information panel with a slightly diagonal yellow divider at the top; dark green footer bar at the bottom.

Text to preserve and typeset clearly:
- Top small heading: {argument name="top heading" default="文明旅游"}
- Main headline in two large yellow-green lines: {argument name="main headline" default="每个人都是一道\n亮丽的风景线"}
- Lower section title: {argument name="section title" default="文明旅游行为公约"}
- Lower pledge items: use exactly 10 items, arranged in two centered rows with green bullet dots: {argument name="pledge items" default="重安全、讲礼仪、不喧哗、杜陋习、守良俗、明事理、爱环境、护古迹、文明行、最得体"}
- Footer text: {argument name="footer text" default="中央文明办｜国家旅游局"}

Typography and color: bold dark green Chinese heading, large yellow headline with green outline/shadow, clean dark green lower text, white panel, yellow accent stripe, dark green footer. Keep the small code “京21110” at the lower right. Do not add extra text, logos, people, borders, or mockup environment.

Create a soft cinematic anime illustration of {argument name="character description" default="a delicate young woman with short tousled pale blue hair and clear blue-gray eyes"} sitting alone on a quiet beach at sunset. She is barefoot on the sand, seated in a curled pose with both knees pulled to her chest and both arms wrapped loosely around her legs, looking gently toward the viewer with a calm, wistful expression. She wears {argument name="outfit" default="an oversized translucent white button-up shirt with rolled sleeves over faded denim shorts"}; the shirt catches the warm backlight and drapes naturally around her body. The scene is a low, intimate three-quarter view near sand level, with the ocean horizon behind her, small waves rolling in, wet sand reflecting the golden sun, soft pastel clouds, and exactly two small distant seabird silhouettes in the sky. Use a warm golden-hour palette with peach, lavender, pale blue, and cream tones, subtle lens flare circles near the lower left foreground, gentle rim lighting on hair and shoulders, delicate linework, painterly shading, realistic beach atmosphere, and an emotional but understated mood. Keep the composition uncluttered, no text, no watermark, no extra characters, no exaggerated details, no overly glossy AI look.

Goal: Create a highly detailed vertical composite photo showing the life progression of {argument name="subject description" default="the same Chinese male"} from birth to adulthood, with six separate age-stage figures arranged in a rising diagonal timeline. Maintain strong facial-structure continuity across all six versions of the same person, while making age, body size, clothing, posture, and hairstyle appropriate for each stage.

Canvas: Vertical 2:3 portrait composition, cinematic studio-photography quality, warm golden light in the upper right fading into cool blue-gray mist in the lower left, subtle starry dust and smoky atmospheric background, soft depth of field, elegant inspirational tone.

Layout: Arrange exactly 6 figures from bottom-left to top-right, increasing in age and height. The youngest baby is at the bottom-left foreground, the toddler stands above and slightly right, the elementary-school child stands behind, the teenager stands higher in the center, the young adult is upper-right center, and the mature adult is tallest at the top-right. Use overlapping bodies and scale changes to create a continuous life staircase.

Subjects: Exactly 6 visible versions of the same male:
1. Newborn, 0 months: sleeping infant wrapped in a pale blue swaddle, lying on a fluffy cream blanket at the bottom-left.
2. Curious toddler, 1–2 years: standing child in a dark blue striped one-piece outfit, small hands raised slightly, short black hair.
3. Happy elementary-school student, 6–8 years: child wearing a navy polo shirt and backpack straps, standing confidently.
4. Teenager, 14–17 years: school-age boy wearing a navy zip-up hoodie with a school crest, white shirt, striped tie, neat short hair.
5. College-life young adult, 20–30 years: young man wearing a crisp white open-collar shirt, modern short black hair, relaxed posture.
6. Confident mature adult, 57 years: tall mature man in a dark business suit, white shirt, tie, pocket square, standing upright and dignified.

Text content: Add exactly 6 translucent rounded-rectangle Chinese labels on the left side, each connected to the matching figure by a thin horizontal line and small dot. Use white serif-style Chinese text. The labels must read, from top to bottom: 「自信的57岁成年人（成人）」, 「大学生络的年轻人（20-30岁）」, 「青少年（14-17岁）」, 「快乐的小学生（6-8岁）」, 「好奇的幼儿（1-2岁）」, 「熟睡的新生儿（0个月）」.

Visual style: Photorealistic composite portrait, high-end commercial retouching, soft rim light, realistic clothing texture, natural skin tones, coherent lighting across all age stages, gentle mist around the lower figures, warm halo around the adult at the top-right.

Constraints: Keep exactly six figures and exactly six matching labels. Do not add extra people, panels, borders, logos, watermarks, or unrelated objects. The faces should be visible and consistent unless intentionally anonymized with {argument name="face treatment" default="plain skin-tone rectangular blur masks over each face"}.

Create a polished anime-style vertical portrait of a melancholic high school student sitting beside a classroom window on a rainy day. The character is {argument name="character gender" default="androgynous male-presenting"}, shown from the chest up in a close three-quarter side view facing left, with their head resting on one hand in a tired, contemplative pose. They have {argument name="hair color" default="dark navy-black hair with subtle purple highlights"}, messy layered bangs, loose wisps around the neck, and the back hair tied into a small tousled bun; one visible ear has a small round silver stud earring. The face should be intentionally covered by a large flat opaque rectangle centered over the facial area, colored muted gray-lavender with a soft vertical gradient, hiding the eyes, nose, and most expression. Dress the character in a refined dark navy school blazer with white piping, brass buttons on the cuff, a crisp white shirt, and a loose {argument name="tie color" default="purple"} necktie. Set the scene inside a quiet classroom with a wooden windowsill, pale curtains, and tall window frames; outside the glass, show rain droplets, a hazy gray-blue city skyline, distant buildings, and blurred trees. Use cool rainy lighting from the window, soft reflections, delicate linework, detailed hair strands, cinematic shading, and a subdued blue-purple color palette. Composition should be vertical 2:3, intimate and slightly cropped at the top of the hair bun and lower torso, with the character occupying most of the frame and the rainy window background visible on the left and right. No text, no watermark, no extra characters.

Goal: Create a wide editorial hero banner about {argument name="headline topic" default="Codex的下一块拼图"}, showing a jigsaw-puzzle metaphor for AI coding, visual design, and missing capabilities.

Canvas: Ultra-wide horizontal banner, approximately 5:2 aspect ratio, dark charcoal technical background with faint grid lines, subtle blueprint marks, tiny dots, and thin circuit-like guide lines. Center the artwork with a large off-white paper/cardboard jigsaw puzzle assembly casting soft shadows.

Layout: Use exactly 15 visible puzzle elements: 13 off-white connected puzzle pieces forming a rectangular mosaic, 1 vivid cobalt-blue puzzle piece inserted at the bottom-right edge, and 1 dashed-outline empty missing-piece slot immediately to the right of the blue piece. The puzzle should look physical, with rounded tabs and sockets, slight bevels, paper texture, seams, grime, and realistic shadowing.

Text content: The main typography dominates the puzzle surface. Use huge bold black sans-serif Chinese type reading {argument name="main title text" default="Codex的"} across the top row and {argument name="subtitle text" default="下一块拼图"} across the middle row. Make the letters oversized, cropped slightly by puzzle seams, and integrated into the tile layout. Keep the text crisp and high contrast.

Visible graphic elements: Include exactly 7 small embedded UI/diagram motifs on puzzle pieces: 1 dark code editor panel in the lower-left with short colored code lines and a small </> symbol; 1 pie chart combined with a small bar chart along the lower middle; 1 simple node-and-connector flow diagram near the lower center-right; 1 bullet-list panel with three round bullets to the right of the flow diagram; 1 user profile card in the upper-right with avatar icon and horizontal text lines; 1 small spreadsheet/dashboard grid in the top-right corner with gray cells and one blue highlighted cell; 1 document/image card on the right side with small mountain thumbnail lines, overlapped by a blue chat bubble containing two white dots and two white horizontal lines.

Visual style: Premium tech editorial illustration, cinematic lighting, tactile paper puzzle texture, muted monochrome palette of black, warm off-white, gray, and accent cobalt blue {argument name="accent color" default="cobalt blue"}. Add subtle scratches, dust, and paper grain. Use realistic shadows between pieces, but keep the whole image clean and readable.

Constraints: No people, no extra logos, no watermark, no random English words beyond tiny unreadable code strokes, and no additional large text. The blue puzzle piece should feel like the key missing component that completes the system, while the dashed empty slot hints that another piece is still absent.

Goal: Create a vertical mobile promotional anime poster for {argument name="app name" default="SynClub"}, themed around {argument name="flower motif" default="roses"} and luxurious gemstones.

Canvas: Tall smartphone aspect ratio, 9:16, full-bleed illustration with a darkened promotional footer overlay at the bottom.

Main subject: A close-up fantasy anime portrait of an elegant young woman from the chest up, centered and facing forward. Her face is mostly covered by a large centered square privacy mask, a soft pastel gradient block shifting from dusty pink on the left to lavender-blue on the right. Visible details include long flowing pastel hair in white, pink, lavender, and pale blue, glossy strands curling around her shoulders, fair skin, and a small visible mouth below the mask.

Costume and accessories: She wears an ornate silver-white jeweled headpiece like a tiara or halo arch with filigree details and blue gemstones. Add exactly 1 rainbow rose tucked near the left side of her hair, with red, yellow, green, blue, and violet petals. Her neckline is covered with elaborate crystal lace jewelry, silver filigree, pearls, and blue teardrop gemstones, including exactly 1 large bright blue teardrop pendant at the center of her chest and exactly 1 smaller diamond-shaped blue pendant hanging beneath it. Her dress is pale, delicate, and bridal, with translucent lace, sparkling highlights, and gem-like embroidery.

Foreground and background: Use a soft dreamy pastel fantasy background with vertical stained-glass-like light streaks in pink, yellow, cyan, and violet. Add exactly 1 red rose in the lower-left foreground, partly overlapping the dress. Keep the illustration highly polished, luminous, romantic, and detailed, with an ethereal manga style and glossy digital painting finish.

Text and promotional overlay: Add faint white Japanese text near the top reading {argument name="top notice text" default="この画像はAIにより生成されるものです"}. At the bottom, place a dark black-to-transparent gradient footer containing exactly 4 promotional elements: 1 red chat-face icon on the far left, 1 white brand wordmark reading {argument name="brand wordmark" default="SynClub"} with “Club” in bold, 1 small Japanese call-to-action reading {argument name="call to action" default="今すぐチャットしよう！"}, and 1 black-and-white QR code in a rounded white square at the bottom right.

Constraints: Preserve the face-obscuring square mask, keep all text legible, use no extra characters, include exactly 2 roses total, and maintain a polished vertical social-media advertisement composition.

Goal: Create an ornate anime fantasy character outfit design sheet for a celestial witch or astral mage, showing a full magical costume and its accessories with a luxurious navy, gold, ivory, and sapphire palette.

Canvas: Wide horizontal fashion reference sheet on warm beige parchment paper, highly detailed 2D illustration, clean line art, delicate gold filigree, elegant gothic-lolita fantasy styling, no realistic photography.

Layout: Arrange exactly 5 main visible elements across the sheet: 1 seated young witch character on the left wearing the full outfit, 1 front-facing full outfit mannequin in the center, 1 circular astrolabe or celestial chart accessory at the upper right, 1 matching pair of thigh-high boots on the far right, and 1 dangling jewel ornament at the lower right. Add small decorative stars, moons, pendants, chains, and constellation motifs floating around the design.

Subject details: The character is a delicate anime girl with {argument name="hair color" default="soft wavy lavender-pink hair"}, pale skin, a calm mysterious expression, and large eyes partly hidden by an enormous pointed witch hat. The outfit theme is {argument name="costume theme" default="celestial star mage"}. She wears an oversized wide-brim witch hat in deep midnight navy, covered in tiny stars, constellation lines, crescent moons, gold chains, dangling star charms, sapphire teardrops, and a large ornate blue gemstone brooch at the hatband. The hat brim is asymmetrical and scalloped, with golden trim and a large crescent-eye motif on one side.

Costume details: The central outfit is a layered robe dress in dark navy and black with ornate gold embroidery, constellation patterns, crescent moons, starbursts, lace-up corset front, high-low layered skirt, ivory ruffled blouse, large white jabot, puffed sleeves, wide cape-like shoulder panels, scalloped hems, hanging tassels, and multiple sapphire jewel medallions. Include white lace cuffs and underlayers with intricate frills and semi-transparent ruffled fabric. Use elaborate metallic gold borders, celestial glyphs, and blue oval gemstones set in sun-shaped frames.

Counted accessory details: Include exactly 4 accessory groups: the huge decorated witch hat, the circular astrolabe chart, the paired thigh-high boots, and the dangling lower-right jewel talisman. The boots are dark navy thigh-highs with black lace tops, gold bands, crescent and star embroidery, pointed high-heel shoe tips, dangling chains, and blue gemstone ornaments. The astrolabe is a round blue-and-gold celestial diagram with concentric rings and radial lines, framed by ornate gold points and small hanging moon and star charms. The talisman is a vertical chain ornament with three blue oval gems, crescent horns, star points, and small dangling drops.

Visual style: Intricate Japanese anime concept art, fashion design sheet, elegant magical-girl gothic couture, fine ink outlines, flat parchment background, deep midnight blue fabric with sparkling star speckles, antique gold metallic accents, sapphire jewel highlights, ivory lace contrast, extremely detailed but clean and readable.

Text: Add a subtle elegant handwritten signature near the bottom right reading {argument name="artist signature" default="Smar"}. Do not add labels, paragraphs, UI elements, or extra readable text.

Constraints: Keep the image focused on the outfit reference sheet, preserve all 5 main elements, emphasize celestial motifs and luxurious embroidery, avoid modern clothing, avoid weapons, avoid cluttered background scenery, and keep the character modestly posed while showcasing the full costume.

Goal: Create a black-and-white manga comic page about {argument name="character name" default="an anonymous young man"} following a fitness lifestyle routine, with all visible faces deliberately obscured by soft rectangular gray blur blocks for privacy.

Canvas: Vertical manga page, approximately 2:3 aspect ratio, clean white margins, crisp black panel borders, grayscale screentone shading, professional Japanese seinen manga line art.

Layout: Use exactly 5 panels arranged as follows: 1 small panel at the top left, 1 wider panel at the top right, 1 large full-width middle panel, and 2 bottom panels with the left panel larger than the narrow right close-up panel.

Panel details:
1. Top-left panel: Close-up of a hand holding a white membership card and inserting or tapping it into a black gym access scanner mounted on a counter or gate. Show the hand, card, scanner slot, and simple gym-entry hardware.
2. Top-right panel: Interior of a modern gym with ceiling lights and fitness machines in the background. Show exactly 2 young adult men: one front-facing man in a fitted plain white T-shirt, and one side/back-facing man in a textured athletic T-shirt. Both have short styled hair and blurred faces.
3. Middle large panel: The main young man exercising on a seated chest-press machine. He wears a fitted athletic T-shirt, sits against a padded backrest, and grips exactly 2 horizontal handles, arms extended outward. Show the machine frame, cables, seat, and weight stack with clean technical linework.
4. Bottom-left panel: Grocery store or supermarket protein section. The main young man is now dressed in a dark business suit, white shirt, and striped tie, holding a shopping basket. The basket contains exactly 4 visible food items: a carton of eggs, a tray of raw meat, a head of broccoli, and a tub labeled "Yogurt". Behind him are refrigerated shelves filled with trays of chicken or meat and small product labels.
5. Bottom-right panel: Close-up crop of the suited man at the supermarket, face blurred, holding or standing beside a carton of eggs. Include a nutrition facts label or product information sheet in the background, drawn with tiny unreadable table lines.

Character details: The main character is a slim, fit young adult man with short layered light hair, clean jawline, and athletic build. Keep his face hidden in every panel using a soft gray rectangular blur, not a black censor bar.

Visual style: Monochrome manga, detailed pen-and-ink contours, subtle screentone gradients, clean indoor perspective, realistic anatomy, polished commercial comic finish, no color except grayscale. Use varied camera angles: close-up, medium gym conversation shot, dynamic workout shot, supermarket medium shot, and tight close-up.

Text content: Only include the visible English word {argument name="yogurt label" default="Yogurt"} on the food tub. Avoid dialogue balloons, sound effects, captions, or extra readable text except small unreadable store labels and nutrition-table lines.

Constraints: Exactly 5 panels, exactly 2 men in the top-right panel, exactly 4 visible grocery items in the basket, all faces blurred, no logos, no watermark, no colored elements, no extra panels.

Goal: Create a dark presentation contact sheet showing six anime cyberpunk character illustration variations generated from the same design, arranged as a 3-column by 2-row gallery on a charcoal background.

Canvas: Square image, approximately 768×850 px, dark gray page background with generous margins. Each artwork tile is vertical portrait format, evenly spaced, with a white caption beneath it. Use a clean sans-serif font, small to medium size, left-aligned under each tile.

Main character: A rebellious anime-style cyberpunk hero named {argument name="character name" default="E1"}, androgynous young adult with shaggy layered {argument name="hair color" default="deep violet hair with golden-yellow underside streaks"}, sharp golden eyes where visible, pale skin, slim athletic build, and a confident dangerous expression. Outfit is a black hooded techwear jacket with yellow hazard-stripe accents, gold trim, triangular insignia patches, black cropped top with yellow emblem, strappy neckline, black cargo pants, belts, gloves, and heavy black boots with yellow laces. Overall palette is black, neon purple, electric violet, and warning-sign yellow.

Gallery layout: Include exactly 6 portrait illustration tiles, each with a distinct composition and caption:
1. Top-left tile caption: “E1 Rooftop Hero”. High-angle rooftop crouching pose above a dense neon city, purple holographic panels floating around, yellow warning triangle icon, dramatic rain-slick ledge and glowing skyscrapers. Add a soft square blur over the face.
2. Top-middle tile caption: “E2 Close-up Broadcast”. Tight close-up portrait from the chest up, hair blowing, purple broadcast interface graphics and city lights behind the character, intense neon rim light. Add a soft square blur over the face.
3. Top-right tile caption: “E3 Full-body Standing”. Full-body character standing on a wet rooftop walkway or fire escape, one hand reaching forward, skyscrapers and purple screens behind, yellow warning signs, glossy reflections on the floor.
4. Bottom-left tile caption: “E4 Livestream Stage”. Front-facing upper-body pose with one hand extended toward the viewer, surrounded by futuristic livestream UI panels, purple hologram screens, small hearts and counters, a bottom banner reading “THIS MESSAGE SHOULD NOT EXIST.” and a small “LIVE” badge.
5. Bottom-middle tile caption: “E5 Seedance Action”. Dynamic action pose, character lunging or sprinting diagonally across a rain-soaked cyberpunk rooftop, motion blur, neon purple city perspective, yellow warning icons, energetic cinematic angle.
6. Bottom-right tile caption: “E6 Softer Standard”. Cleaner softer portrait on a pale lavender tech-interface background, bust-up framing, same jacket and outfit, calmer lighting, minimal neon UI elements. Add a soft square blur over the face.

Visual style: Highly polished anime illustration, cyberpunk noir atmosphere, dramatic rim lighting, glossy wet surfaces, dense futuristic city detail, purple holograms, yellow hazard iconography, sharp linework, rich contrast, cinematic composition. The six tiles should look like consistent variations of the same character and world.

Constraints: Use exactly six tiles and exactly the six captions listed above. Preserve the contact-sheet layout and dark background. Keep the labels readable in English. Do not add extra panels, watermarks, logos, or unrelated characters. Use the background city theme {argument name="setting" default="rainy neon cyberpunk metropolis"} and overall mood {argument name="mood" default="dangerous heroic hacker-broadcast energy"}.

Create a tall vertical Korean manhwa / moe fantasy illustration of a mysterious young woman standing in front of a glass window or translucent panels overlooking a calm ocean at sunset. The composition is a close full-body portrait cropped tightly, with the woman centered and facing forward, her face mostly in shadow and expression unreadable. She has long windblown dark hair with pale silver-gray ends, blunt bangs covering her eyes, soft pale skin, and a slender elegant figure. Her body is stylized like living rose-themed glass or shadow: glossy translucent skin, dark botanical silhouettes across her torso and legs, scattered petal reflections, and a strapless dark green moss-like rose bodice made of organic foliage. Add dramatic rim lighting from the right side, with a bright sun flare near her chest and shoulder, cool blue shadows, and pink-orange sunset highlights. Surround her with exactly 9 visible red roses: 1 large partial rose at the upper left corner with leaves, 1 upright rose near the upper right beside her head, 1 large rose at lower center-left in front of her hip, 1 rose bud/half-open rose at lower center-right, 1 full rose at lower right-center, 1 small rose near the lower right, 1 large partial rose cropped at the bottom left corner, 1 pale cream rose cropped at the far lower right edge, and 1 small rosebud on the right edge near the horizon. Include floating red petals and tiny glowing specks around her, plus vertical glass-panel seams and reflections. Background: distant cliffs and islands on the left, turquoise-blue sea, pastel sky at dawn or dusk, soft atmospheric haze. Use highly polished anime concept art, Korean webtoon beauty, semi-realistic lighting, delicate linework, glossy reflections, rich crimson roses, cinematic backlight, dreamy romantic mood, no text, no watermark. Character customization: {argument name="character name" default="unnamed rose spirit"}, {argument name="hair color" default="black fading to silver-gray"}, {argument name="dress material" default="dark green moss-like rose foliage"}, {argument name="flower type" default="red roses"}, {argument name="background setting" default="ocean sunset seen through glass panels"}.

Create a dark gothic fine-art portrait in a vertical 9:16 canvas: one pale adult woman lies supine like an ornate funeral tableau, surrounded by a dense, nearly continuous carpet of deep burgundy-black roses and dark foliage. She wears an elaborate {argument name="dress color" default="black"} Victorian mourning gown with high neck, long lace sleeves, embroidered floral appliqué, sheer panels, layered satin and tulle, and intricate beadwork. Her long {argument name="hair color" default="black"} hair spreads around her head; one ear with delicate earrings is visible, while her face is intentionally covered by a centered soft-edged square blur/censor block in muted brown tones. Pose: one hand rests on her torso, the other arm extends down to the side, body angled slightly diagonally from upper center to lower center. Surround her with three clearly visible antique objects among the roses: one dark ornate book with a metal cross emblem at the lower right, one gilded picture-frame edge at the left side, and one small metallic crucifix or decorative cross near the roses. Use a moody {argument name="lighting style" default="low-key chiaroscuro"} lighting setup with very deep shadows, subtle highlights on lace, skin, and rose petals, a muted palette of black, oxblood, antique gold, and charcoal, hyper-detailed texture, romantic gothic opulence, cinematic realism, top-down composition, solemn atmosphere, no text, no watermark.

Create a tall vertical surreal anime-style illustration of {argument name="character name" default="a silent young woman"} shown in elegant side-profile silhouette at the center, facing right, with a slim graceful body and a short bob haircut. The figure is a dark translucent double-exposure silhouette filled with a serene ocean coastline, rocky islands, reflective water, and soft dusk clouds; her head and torso blend seamlessly with the seascape inside her body. Place pale blue-white roses tucked into the hair near the back of the head, with subtle lace-like floral detail near the shoulder. The background is a calm endless sea at sunset, horizon line low-mid frame, glowing peach and gold light near the right side, fading into cool cyan and deep teal shadows. Add {argument name="flower type" default="deep crimson roses"} floating around the composition: exactly 8 visible red roses total — 3 large cropped roses along the bottom edge, 2 small side roses at mid-left and mid-right, and 3 roses near the top edge — with dark green stems and leaves where visible. Use a dreamy Korean webtoon / moe illustration style mixed with painterly fantasy concept art, soft atmospheric haze, reflective water texture, high contrast between the dark silhouette and luminous sky, romantic melancholy mood, no text, no watermark, no border, portrait 9:16 composition.

Create a high-detail anime cyberpunk key visual of {argument name="character name" default="Tomoshibi Kii, short-haired version"}: a fierce young woman crouching on the narrow edge of a skyscraper rooftop high above a dense futuristic city at night. Use a dramatic vertical 2:3 composition, extreme high-angle perspective looking down into glowing canyons of buildings, with the character dominating the foreground and centered slightly right. She has short tousled {argument name="hair color" default="deep violet hair with yellow-gold underlayers and streaks"}, sharp golden eyes, pale skin, a serious defiant expression, and a small black headset microphone at her cheek. Outfit: black tactical streetwear with gold accents, including a high-collar oversized hooded jacket, cropped black top with strap details, choker, fingerless gloves, loose cargo pants, knee padding, heavy boots, belts, zippers, brass snaps, and a shield-like geometric insignia patch on the sleeve. Lighting is moody neon purple and amber, with strong rim light on hair and jacket edges, glossy fabric highlights, rain-damp rooftop reflections, and dense detail in the city windows. Background: towering skyscrapers packed with tiny lit windows, glowing purple holographic billboards, orange warning UI graphics, and a vast night skyline fading into darkness. Include exactly 4 prominent holographic screens: 1 large purple portrait screen of the same girl on the upper left, 1 smaller purple emblem/abstract symbol screen on the lower left, 1 orange warning triangle screen on the upper right, and 1 purple waveform/data screen on the mid-right. Style should be polished Japanese anime key art, intricate linework, cinematic depth, luminous cyberpunk atmosphere, ultra-detailed background, sharp focus on the character, no visible watermark, no extra readable text beyond abstract UI marks.

Goal: Create a nostalgic Japanese height-difference illustration titled {argument name="headline text" default="大切な家族"}, showing a beloved family pairing measured on a vintage height chart: a schoolgirl and a tiny calico kitten facing each other with a cute emotional gap-moe feeling.

Canvas: Vertical portrait poster, warm cream parchment background, soft sepia ink linework, delicate watercolor shading, no harsh shadows. Use an elegant antique stationery style with thin brown decorative borders, small flourishes, and a faint “AI” mark in the upper-left corner.

Layout: Place a tall central measuring ruler from bottom to near top, framed by a narrow ornate border. The ruler is labeled vertically in Japanese as 「身長計」 at the top, with centimeter marks from 0cm to 200cm; show major labels every 10cm and include visible labels around 160cm, 165cm, and 170cm. Add two dotted horizontal guide lines: one at 165cm extending left to the girl’s head, and one at 20cm extending right to the kitten.

Subjects: On the left of the ruler, draw {argument name="tall character" default="a gentle Japanese schoolgirl"} in side profile facing right. She is exactly 165cm tall, standing upright with a shy, affectionate posture, long dark brown hair, blurred/featureless face, cream sailor-style school blouse with dark collar and cuffs, brown necktie, black pleated skirt, black knee-high socks, and brown loafers. On the right of the ruler near the bottom, draw {argument name="small character" default="a fluffy calico kitten"} at exactly 20cm tall, facing left and looking up, with white, orange, and dark patches, round body, short legs, and curled raised tail.

Information cards: Include exactly 2 ornate framed profile cards. Left card beside the girl: label “Name”, value {argument name="tall character name" default="結衣"}, label “Height”, value “165cm”. Right card beside the kitten: label “Name”, value {argument name="small character name" default="むぎ"}, label “Height”, value “20cm”. Use brown serif English labels and Japanese name text inside inset boxes.

Speech/thought bubbles: Include exactly 2 cloud-shaped bubbles. Left bubble near the girl reads 「今日もかわいいな」 with two small thought circles leading toward her. Right bubble above the kitten reads 「おやつまだニャ？」 with two small thought circles leading toward the kitten.

Visual style: Soft anime-inspired storybook illustration, clean fine outlines, muted beige and brown palette with subtle black uniform accents and natural calico colors. Keep the composition symmetrical around the measuring chart, with generous empty space and a calm sentimental mood.

Constraints: Preserve all visible Japanese text exactly as written, keep the scale relationship clear and humorous, include only the two characters and the two profile cards, avoid photorealism, avoid extra animals or objects, and do not add modern UI elements.