GPT Image 2 prompts

Draw a {argument name="subject" default="beautiful girl"} in {argument name="style" default="Japanese anime style"}

Generate a {argument name="composition" default="9:16 vertical"}, high-precision ancient Chinese fantasy-style refined portrait of a woman. The main subject is a young adult Eastern woman, approximately {argument name="age" default="20–28"} years old.

Create a cinematic hyper-realistic portrait of a {argument name="character age and gender" default="21-year-old woman"} sitting casually on weathered stone stairs in a quiet park at twilight, beside an old glowing lamp post. She has a soft natural smile, relaxed posture, arms resting on her knees, and slightly messy dark hair styled in a long side braid. Outfit includes exactly 5 visible pieces: a {argument name="shirt color and style" default="pink heavyweight oversized T-shirt"}, black cargo pants, white sneakers, a black smartwatch on one wrist, and minimal natural makeup. Compose her seated on the right side of the frame, with the vintage lamp post on the left emitting warm yellow light; add two small blurred park lights in the background. Use shallow depth of field, soft bokeh trees, warm dusk ambience, realistic skin texture, natural proportions, and cinematic color grading. The stone staircase should look aged, stained, and textured, with a curved side wall leading diagonally through the composition. Shot on a full-frame camera with an 85mm lens look, f/1.8 depth of field, balanced twilight exposure, ultra-detailed photorealism, no text, no watermark.

Create a warm, whimsical photorealistic miniature scene in a cozy pastel study workspace: a tiny cute young South Asian woman, {argument name="character name" default="unnamed girl"}, sits on top of a giant modern smartphone lying flat on a desk, as if the phone is a glossy black platform. She has long wavy dark hair with a side part, delicate drop earrings, and wears an oversized chunky-knit {argument name="sweater color" default="soft blush pink"} sweater dress, white socks, and chunky white sneakers; pose her seated casually with one leg crossed over the other and both hands resting beside her on the phone. Use a dreamy golden-hour aesthetic with sunlight streaming through a large window on the left, soft haze, bokeh, shallow depth of field, and a feminine pink-beige color palette. Surround her with exactly 11 visible workspace objects: 1 rose-gold giant smartphone, 1 open laptop on the right with a pink screen showing motivational script text, 1 pink coffee mug with latte art and a gold handle, 1 pair of pink over-ear headphones in the foreground right, 1 pink spiral notebook in the foreground with gold handwritten words “ideas plans action”, 1 pink pen, 1 pink pencil cup holding pens and pencils, 1 small succulent in a ribbed pot, 1 small desk sign reading “good things take time”, 1 glass vase of airy pale pink flowers behind the girl, and 1 small decorative tray with pastel accessories in the foreground left. Background details include soft shelves with books and pink decor, sheer curtains, and an elegant desk surface. Make the phone enormous relative to the miniature woman, preserve realistic shadows and reflections on the black screen, and keep the overall mood cozy, aesthetic, luxury, cute, and social-media-ready. Avoid extra people, extra text, logos, watermarks, or distorted anatomy.

Goal: Create a polished vertical promotional poster for {argument name="product name" default="GPT Image 2"} explaining that the image model thinks before rendering, using a playful 3D toy-like visual style.

Canvas: 4:5 vertical social media poster, deep cobalt-blue background with a subtle radial glow and soft vignette, high-resolution glossy 3D render aesthetic.

Main visual: Center the top half around a surreal 3D object: a rounded glossy royal-blue cube fused with a shiny pink brain hemisphere protruding from the cube's right side. The cube is angled slightly so the front-left face and top are visible, with soft studio highlights, rounded edges, and a grounded shadow. The brain should look like inflated candy or silicone, with rounded folds, bright pink highlights, and a soft magenta glow where it meets the cube.

Decorative elements: Scatter exactly 34 small floating candy/confetti objects around the poster, avoiding the main text area. Include these visible types: 1 yellow star, 2 curved jellybean shapes, 11 short capsule sprinkles, 11 small spheres/dots, 4 larger glossy balls, 3 tiny round dots, and 2 partial edge sprinkles. Use pink, yellow, mint green, purple, orange, white, and red, with soft shadows and glossy highlights.

Text layout: Use bold white sans-serif typography, centered. Large headline in the middle: “{argument name="headline text" default="It reasons, then renders."}” on two lines, very heavy weight, clean geometric sans serif, slight shadow for contrast. Beneath it, smaller bold subtitle: “{argument name="subtitle text" default="GPT Image 2 plans your image before it makes a pixel."}” on two centered lines.

Instruction card: At the lower third, place a rounded off-white command-card panel with a subtle drop shadow, like a copyable prompt box. On the right side, attach a coral-red rounded rectangular button partially behind the card, labeled “COPY” in dark bold text. Inside the card, show exactly 4 monospaced lines in dark charcoal: “PLAN: composition and constraints.”, “SEARCH: facts you need, or SKIP.”, “GENERATE: the brief.”, “VERIFY: check output before showing.”

Footer: Bottom-left, add a tiny white line-art dog/mascot icon followed by the text “{argument name="footer label" default="GPT IMAGE 2 · PLAIN AND SIMPLE"}” in small uppercase white sans serif. Bottom-right, add the progress text “{argument name="progress text" default="1 / 7"}”.

Visual style: Bright modern 3D advertising, glossy plastic materials, candy colors, rounded friendly forms, strong contrast, crisp typography, soft realistic shadows, no photorealistic people, no watermark, no extra text.

Goal: Create a horizontal social-media art poster combining a vintage French melodrama painting with a black poetry-and-credits panel.

Canvas: Wide 16:9 composition, approximately 1024×576. The image is divided into two main columns: a large painted scene on the left taking about 60% of the width, and a flat black text panel on the right taking about 40%. Add a thin dark brown border around the full artwork.

Left scene: Paint an early 20th-century European interior in a muted post-impressionist / Toulouse-Lautrec-inspired oil-painting style with visible brush texture, aged varnish colors, ochre, teal, cream, and black. Show exactly 2 women in a tense dramatic conversation. Woman 1 stands on the left in a loose white blouse and dark skirt, turned toward the other woman, both hands raised in alarm with spread fingers. Woman 2 sits on the right on a bed or cushioned bench, wearing a low-cut black slip dress and black stockings, posed coolly with one elbow bent and hand near her chin, one knee raised, giving the impression of a dangerous femme fatale. Behind them include exactly 3 major background elements: a tall dark wooden mirror, a narrow vertical wooden panel or door, and an iron bedframe. Both women's faces should be obscured by soft rectangular blur blocks, as if anonymized or censored.

Subtitle: At the bottom of the painted scene, add a white italic film subtitle with a thin black shadow reading “{argument name="French subtitle" default="Vous êtes folle!—"}”. Place it centered near the bottom edge of the painting.

Right panel layout: Solid black background. At the top center place one small beige square QR-code-like graphic with exactly 3 visible black corner markers. Below it, center a poem in hand-lettered golden ochre text. Use casual uneven handwritten lettering, not a standard clean font.

Main poem text: “{argument name="poem text" default="you’re insane!\nshouts the distraught woman\nin French—\nI thank goodness for subtitles\nand dangerous femme fatales"}”. Keep the lines centered and stacked, with generous spacing.

Credits text: In the lower middle of the right panel, add smaller golden handwritten text reading: “AI image created by\n{argument name="artist name" default="Michael H. Lester"}\nin collaboration with\nGPT Image 2 Low on\nNightcafestudio\ntwitter: mlhlester”.

Bottom links: Along the bottom of the right panel add two centered golden lines: “{argument name="creator link" default="https://creator.nightcafe.studio/?ru/mhlester"}” and “{argument name="portfolio link" default="https://4-michael-lester.pixels.com/"}”.

Constraints: Use exactly 2 women, exactly 1 subtitle line, exactly 1 QR-code-like square, exactly 1 main poem block, exactly 1 credits block, and exactly 2 bottom link lines. Keep the right panel mostly empty black space, maintain a moody noir theatrical tone, and avoid photorealism.

Create a realistic paparazzi-style street photograph of {argument name="character name" default="an elegant woman"} stepping out of the Four Seasons hotel entrance in a busy city. She is centered full-body, walking confidently from the curb toward the camera, wearing a fitted black short-sleeve wrap dress with a deep V neckline and a high thigh slit, black pointed-toe stiletto pumps, and carrying a small structured black handbag in her left hand. Her face is intentionally anonymized with a soft rectangular blur, while her medium brown shoulder-length hair is visible in natural sunlight. Surround her with exactly 3 hotel security or doorman figures on the left near the doorway and a dense crowd of photographers on the right holding cameras with flashes aimed at her. Include a glossy black luxury sedan parked at the curb on the right foreground with a partial European-style license plate. The background should show the Four Seasons entrance sign, stone architecture, doorway shadows, and bright midday sunlight casting strong realistic shadows on the pavement. Style: high-end celebrity candid photo, editorial realism, 35mm street photography, natural contrast, slight motion energy, no glamour retouching, vertical 3:4 composition.

Create an ultra-realistic close-up semi-portrait of {argument name="character description" default="an exceptionally handsome Black man in his late 20s"}, photographed from the chest up and facing directly toward the camera inside an architecturally stunning luxury residence on a bright sunny afternoon. He has rich dark brown skin, very short tightly coiled black hair visible at the top edge, a strong neck and collarbones, and one ear visible on the right side. His face is completely covered by one large centered opaque square censor block in a dark warm brown color, leaving only the top of the hair, right ear, lower chin/neck area, and clothing visible. He wears {argument name="outfit" default="a dark charcoal tailored blazer over an open-collar dark shirt"}. The background is a modern upscale interior with tall black-framed windows, warm wood tones, soft golden sunlight, and shallow depth of field with creamy bokeh. Use cinematic natural light, realistic skin texture on the visible neck, high dynamic range, 85mm portrait lens look, crisp foreground detail, blurred architectural background, vertical 3:4 composition, no text, no watermark.

Create a cinematic, ultra-realistic nature landscape at magical sunrise: a pristine alpine valley with a shallow winding river in the foreground, framed by dense evergreen pine forests and jagged snow-dusted mountain peaks in the background. The sun is low on the left horizon, casting warm golden light, long rays, glowing mist, and soft volumetric fog through the trees and across the valley. In the river, show exactly 4 spotted deer drinking or grazing at the water’s edge: three grouped near the center with heads lowered, and one smaller deer farther right near the riverbank. The foreground is filled with vibrant wildflowers, especially purple lupines and small yellow blossoms, with lush green grasses, stones, and river rocks. Use {argument name="time of day" default="golden sunrise"}, {argument name="main animals" default="four spotted deer"}, {argument name="mountain style" default="jagged alpine peaks with light snow"}, {argument name="flower colors" default="purple and yellow wildflowers"}, and {argument name="mood" default="magical, peaceful, cinematic"}. Make the composition wide and immersive, with strong depth from flowers in the foreground to river, forest, misty valley, and towering mountains. Photorealistic detail, natural colors, dramatic warm lighting, soft haze, high dynamic range, serene wilderness atmosphere, no text, no people, no buildings, no watermark.

Create a vertical cinematic close-up portrait of {argument name="character name" default="a graceful female immortal cultivator"} in a Chinese xianxia fantasy drama style. She is framed from the chest up in a dim antique interior beside a softly glowing lattice window, with hazy volumetric sunlight streaming from the upper left and a dark blurred background. The subject has long glossy {argument name="hair color" default="black"} hair styled half-up with loose flowing strands over the shoulders, an ornate gold hairpin visible near the back of the head, and delicate period-drama styling. Her face is deliberately covered by one large opaque muted gray-brown rectangular censor block centered over the entire face, while the hair, neck, clothing, and hands remain visible. She wears layered flowing Hanfu robes in pale ivory, silver-gray, and light blue, with embroidered brocade trim, a warm red inner collar, soft translucent fabric, and subtle metallic thread details. Both hands hold a vivid translucent emerald-green sword or jade blade horizontally across the lower foreground, angled slightly upward from left to right, with gold fittings at the right edge; the blade catches a bright specular highlight. Use realistic high-budget Chinese fantasy TV production lighting, shallow depth of field, soft bloom, film grain, natural skin and fabric texture, dramatic backlight, and a mystical but grounded mood. Vertical 9:16 composition, close crop, no visible text, no watermark, no extra characters.

Create a vertical anime-style summer street illustration of a young office woman stopping outside a traditional Japanese noodle shop on a blazing hot day. The woman, {argument name="character name" default="a tired office lady"}, is in the right half of the frame, bent forward with one hand on her knee and the other holding a structured black handbag, visibly exhausted and sweating. She has long wavy {argument name="hair color" default="black hair"}, flushed skin, sweat droplets on her face, neck, arms, and legs, and wears a navy business blazer and matching short skirt over a white blouse, black loafers, and a small pink portable neck fan hanging from a strap. Her face should be intentionally obscured by a centered soft-edged square blur/mosaic censor block. On the left foreground, place a wooden A-frame restaurant signboard angled toward the viewer with a colorful illustration of hiyashi chuka: one bowl of chilled noodles topped with exactly five visible toppings—green cucumber strips, pink ham strips, yellow shredded egg, red pickled ginger, and one red tomato slice. The signboard text should be large vertical Japanese calligraphy: {argument name="sign text" default="冷やし中華 始めました"}. Behind it, show the restaurant entrance with dark blue noren curtains carrying large pale characters reading {argument name="shop curtain text" default="初夏屋"}, plus a narrow side banner with handwritten soba-style characters. Add comic heat effects: exactly two red handwritten exclamations, one near the upper right reading {argument name="heat exclamation" default="あづ〜〜〜…"} and one near the left of the woman reading "はぁ〜", exactly four small white puff clouds around her head and shoulders, and multiple wavy white heat lines rising from the pavement and storefront. The background is a sunlit old shopping street with wooden eaves, stone pavement, bright blue sky, distant buildings, harsh highlights, and shimmering heat haze. Use detailed cel-shaded anime rendering, dynamic perspective, warm saturated sunlight, crisp line art, glossy sweat highlights, and no extra characters, no watermark, no English text.

Create an ultra-detailed dark fantasy alchemist workshop scene at night, vertical 2:3 composition. Inside a cramped medieval stone-and-timber laboratory filled with hanging dried herbs, brass astrolabes, glassware, old books, crystals, and copper apparatus, show exactly two female alchemists: a younger apprentice on the left with dark curly hair, layered gray-brown robes, belts, pouches, bracelets, and necklaces, carefully pouring luminous liquid from a small glass vial into a round-bottom flask; and an older mentor on the right with {argument name="mentor hair color" default="silver-gray hair"} tied up, wearing dark layered robes and jewelry, holding an open ancient book while gesturing as if instructing. Both faces are intentionally covered by plain soft-edged brown square privacy masks, with no facial features visible. The central workbench contains exactly three major glowing vessels: one large round flask glowing pale green at center, one pear-shaped flask glowing blue on the right, and one small blue vial near the furnace. Include one open illustrated botanical grimoire in the lower left, one black mortar and pestle, one ornate circular brass plate, one shallow bowl of stones, and scattered crystals in blue, purple, clear, and smoky tones. On the right, include one warm orange furnace or alchemical boiler with copper pipes and gauges. In the background, show one arched window opening to a moonlit pine forest with exactly two visible moons in the blue night sky. Lighting should mix cold moonlight from the left window with warm firelight from the right furnace and magical cyan-green glow from the liquids. Use intricate painterly realism, highly textured wood, stone, aged metal, glass reflections, botanical clutter, gothic witchcraft atmosphere, cinematic depth, and no modern objects or readable text.

Create a warm, high-detail anime illustration of two maid sisters hugging in a sunlit room. The scene shows exactly 2 characters: an older sister on the right, taller, wearing a classic black long-sleeve maid dress with a white frilled apron, white ruffled maid headband, black ribbon bow at the collar, and long {argument name="older sister hair color" default="warm brown"} hair tied into a high side ponytail with a black bow; and a younger sister on the left, shorter, joyfully hugging her, wearing a matching black-and-white maid outfit with a frilled apron, black bow tie, white headband, and {argument name="younger sister hair color" default="light blonde"} twin ponytails tied with small black bows. The younger sister has closed smiling eyes, rosy cheeks, and an open happy smile while pressing her cheek against the older sister’s chest; the older sister is embracing her gently, with her face intentionally covered by one centered solid rectangular censor block in a muted skin-tone brown. Use a cozy Victorian-inspired interior with a large bright window on the left, sheer white curtains, golden afternoon sunlight streaming diagonally through dust motes, soft bloom, warm beige walls, a framed picture on the right wall, and a small wooden cabinet with a vase of white flowers. Composition is vertical portrait, waist-up to knee-length framing, affectionate sisterly pose, soft rim lighting, delicate fabric folds, frills, lace, ribbons, glossy hair highlights, dreamy atmosphere, polished modern anime style, no text, no watermark, no extra characters.

Create a square, cinematic close-up portrait of {argument name="subject" default="an androgynous young adult with tousled dark hair"} in a dim indoor setting, cropped tightly so the head and shoulders fill the frame. The person wears a {argument name="clothing" default="dark charcoal knit sweater"} and rests one hand near the chin in a thoughtful pose, with softly visible fingers at the lower center. Use moody low-key lighting, shallow depth of field, muted warm-gray background, realistic skin and hair texture, and a subdued film-photography look with slight grain. Place a large opaque square privacy block directly over the center of the face, covering most facial features from the forehead down to the mouth; the block should be a smooth vertical gradient from near-black dark brown at the top to medium warm brown at the bottom, with sharp straight edges, centered and occupying most of the image width and height. Keep only the hair edges, one ear area, shoulder, sweater, and hand partially visible around the block. No text, no watermark, no decorative elements.

Using the provided reference image as the character and car base, transform it into a lively hand-drawn Japanese anime character-reference collage on a white sketchbook-style background. Keep the same person, black outfit, camera, white open-top car with red seats, and round roadside sign recognizable, but reinterpret everything with expressive manga linework, soft gray shading, cute proportions, sparkles, hearts, musical notes, and handwritten Japanese annotation doodles.

Layout: Create exactly 11 character/car vignettes around the page:
1. A large upper-left bust portrait wearing round black sunglasses.
2. A small top-center drawing of the white convertible driving near the road sign.
3. A large upper-right excited smiling head portrait.
4. The main central full-body standing pose in front of the car, holding the camera.
5. A mid-left seated back-view scene looking toward the water/landscape with the camera beside them.
6. A small left chibi holding the camera up near their face.
7. A lower-left chibi driving the white convertible.
8. A lower-center bust portrait looking over the shoulder.
9. A tiny bottom-center chibi standing still with sunglasses.
10. A right-middle chibi seated in a folding chair beside the car.
11. A lower-right walking side-view pose holding the camera, with motion lines.

Additions: Fill the empty space with casual handwritten Japanese notes, arrows, checkboxes, tiny hearts, stars, birds, clouds, music notes, and energetic sound-effect marks. Include one checklist card on the right titled in Japanese, styled like a travel packing memo. Add small decorative labels such as {argument name="main handwritten title" default="ちびロドちゃん"}, {argument name="top praise text" default="最高〜っ!!!"}, and short travel-themed notes about driving, coffee, sunglasses, and taking photos.

Style: Cute monochrome-and-muted-color manga sketch, loose black ink outlines, pale gray shadows, limited color accents matching the reference, playful diary-page composition, dense but readable spacing. Preserve the reference character’s general hairstyle, clothing silhouette, camera, car color, red seats, and sign motif. Do not make it photorealistic; make it look like a polished illustrated character sheet full of doodles.

Using REFERENCE_0 and REFERENCE_1 as the base design for the small StackChan wheeled robot, transform it into a cheerful retro Japanese manga/anime concept sheet for a cute “friend from the future.” Keep the recognizable white cube head, black screen face, side StackChan label, gray body, and black/yellow mecanum wheels, but stylize everything as hand-drawn 1970s–1980s children’s manga with ink outlines, halftone texture, warm cream paper, and slightly faded print colors.

Goal: Create a full-page character introduction poster for {argument name="character name" default="StackChan"}, presented as a futuristic robot-cat companion who is good at turning “?” into “!”. Add cat ears, a cat tail, a friendly animated face on the screen, and playful manga energy.

Layout: Use a vintage Japanese magazine/comic-page layout with exactly 8 main content areas:
1. Large hero illustration on the left: the robot-cat waving happily, with the title above.
2. Top-right action panel: the robot moving fast on its mecanum wheels with speed lines and the sound effect 「ビューン」.
3. Middle-right feature panel: the robot projecting a blue hologram information display, with a speech bubble showing 「！」.
4. Expression strip: exactly 5 small face-expression icons labeled in Japanese — happy, yay, confused with a question mark, surprised, and sleepy.
5. Lower-left profile box with bullet-point facts.
6. Bottom-center cameo panel: two famous blue robotic-cat and schoolboy-inspired silhouettes with black censor bars over their eyes, standing beside StackChan, with a speech bubble saying 「スタックチャン、すごいね！」.
7. Lower-right secret gadget card featuring a ring-shaped device.
8. Bottom-right concept message box explaining that the character turns everyone’s 「？」 into 「！」.

Text: Include Japanese typography and furigana-style small kana for a retro children’s encyclopedia feel. Main headline should read {argument name="headline text" default="22世紀から来た未来の友だち"}. Big red title should read {argument name="title text" default="スタックチャン"}. Feature captions should describe free movement by mecanum wheels, projecting information with holograms, and the gadget 「未来リンク」 that connects to the future network.

Style constraints: Make it look like a scanned vintage manga character guide page, not a photo. Use bold red title lettering, cream background, blue/pink/yellow feature panels, thick black panel borders, hand-lettered Japanese captions, visible print grain, and whimsical child-friendly sci-fi energy. Do not add more than the 5 expression icons or more than the single secret gadget card.

Goal: Create a minimalist Swiss-style exhibition poster for a generative nature bonsai show, centered on a realistic bonsai tree in a ceramic pot, combining botanical photography with precise algorithmic diagram graphics.

Canvas: Vertical poster, 3:4 aspect ratio, off-white warm paper background with generous negative space, crisp print-design finish.

Main subject: A single mature bonsai tree, photorealistic, placed slightly below center. The tree has a thick twisted dark brown trunk, exposed roots, dense bright green pine-like foliage arranged in layered horizontal pads, moss-covered soil, and a shallow round gray concrete ceramic pot with subtle stains and shadow. The bonsai should look like a premium product photograph integrated into a design poster.

Typography: Use thin modern sans-serif uppercase typography, black and dark gray only. The main headline at upper left reads “{argument name="headline text" default="BONSAI SYSTEM"}” in very large thin letters over two lines. Above it, small text reads “GENERATIVE NATURE” and “SERIES 01”. Below the headline, write “GENERATIVE FORM”, then a small three-line list: “ALGORITHM”, “VARIATION”, “EQUILIBRIUM”. On the upper right, write “{argument name="city and year" default="TOKYO 2026"}”. On the lower left, write the event dates as three stacked lines: “{argument name="event dates" default="MAY 08 / JUN 21 / 2026"}”, with 2026 much larger. Under that, add “AT” and “TOKYO GENERATIVE LAB”. Near the left middle, add the small slogan “NATURE IS THE ORIGINAL ALGORITHM.” Add vertical Japanese text beside it: “盆栽系統”. At bottom center, include exactly 5 tiny numbered index labels: “01 SEED”, “02 GROWTH”, “03 FORM”, “04 BALANCE”, “05 TIME”. At bottom right, include “WWW.GENERATIVE-LAB.JP”.

Geometric layout: Overlay a delicate technical diagram system in very light gray and black around the bonsai. Include exactly 1 large faint circle centered behind the tree, exactly 1 vertical centerline running from top to bottom, exactly 1 horizontal line crossing the composition through the foliage area, exactly 2 small black circular nodes on the horizontal line, exactly 1 rectangular grid panel in the upper right behind the foliage, exactly 1 fan of many thin curved contour lines sweeping down from the upper right, exactly 1 smaller nested curved-line corner motif near the lower right, exactly 2 dotted matrix blocks made of tiny black dots, exactly 2 small pale gray square blocks, and exactly 1 narrow solid black vertical rectangle near the upper-right grid.

Layout details: The poster should feel mathematically composed and editorial, with the bonsai occupying the center-lower area and the typography aligned to strict margins. Use thin divider strokes and short horizontal rules as accents. Maintain a clean monochrome graphic system with the only strong color being the green bonsai foliage.

Visual style: High-end museum poster, Japanese design influence, Bauhaus/Suisse typography, generative algorithm diagram aesthetics, precise vector overlays, subtle paper texture, photorealistic plant and pot, no clutter.

Constraints: Keep all text sharp and legible, do not add extra plants or people, do not add a watermark, use exactly one bonsai tree and one pot, preserve the sparse off-white background and black geometric design language.

Create a minimalist black-and-white conceptual diagram on a wide 16:9 white canvas showing a diffusion-transformer creature parasitizing an autoencoder latent space. Use thin hand-drawn black outlines and simple sans-serif labels, with a clean article-illustration style. Across the lower third, draw a horizontal pipeline: a long white tube enters from the left with a dashed centerline, passes through a small rounded arch labeled “Encoder,” then becomes a solid black rectangular central block labeled “Latent” in white, then exits through a matching rounded arch labeled “Decoder” into a white tube on the right with a dashed centerline. Above the latent block, draw one very large oval speech-bubble or callout balloon with a pointed tail aimed down toward the latent block. Inside the balloon, centered slightly left, draw exactly one small blob-like creature: soft amoeba body, two tiny dot eyes, four rounded drippy legs, and one raised antenna ending in a small circle; make it mildly uncanny but not scary and not overly cute. To the right of the creature inside the balloon, place the label “{argument name="creature label" default="DiT"}” in black. Use exactly three visible text labels on the diagram: “Encoder,” “Latent,” and “Decoder,” plus the creature label. Keep the composition sparse, monochrome, no shading, no color, no extra icons, no watermark.

Goal: Create a high-energy 12-panel anime storyboard collage for an “AI super-dimensional soccer” scene, blending magical-girl idol aesthetics with futuristic sports action. The story follows {argument name="character name" default="three flower-themed anime soccer girls"} dribbling and passing through an AI-powered stadium, ending in a cute heart-shaped climax.

Canvas: Wide horizontal collage, 4:3 aspect ratio, divided by thin white gutters into exactly 12 numbered panels. Use bold white panel numbers in the upper-left corner of each panel, numbered 1 through 12. Overall palette: electric violet, neon blue, cyan, magenta, glowing pink hearts, and dark stadium shadows.

Layout: Arrange the panels in a manga/storyboard grid with varied panel sizes. The top row contains panels 1, 2, and 3. The second row contains panels 4, 5, and 6. The third row contains panels 7, 8, and 9, with panel 9 split internally like a small reaction collage. The bottom row contains panels 10, 11, and 12.

Subject details: Use exactly 3 main anime soccer girls: one silver-blue-haired girl, one dark-haired girl with purple flower hair ornaments, and one lavender-pink-haired girl. They wear matching ornate futuristic soccer-idol uniforms: white and navy frilled tops, short layered skirts, thigh-high socks, boots, metallic trim, small floral accessories, and glowing circuit-like accents. Include one purple-and-white soccer ball with star markings. Add one cute humanoid AI robot mascot in the audience/reaction panel, with a dark glowing face and heart-shaped eyes.

Panel-by-panel content, exactly 12 numbered scenes:
1. Silver-blue-haired girl sprints and dribbles the star soccer ball at extreme speed, leaving icy blue digital motion trails and splashing energy across the field.
2. Close-up of the dark-haired strategist girl beside a transparent purple holographic tactics board filled with soccer-field diagrams, data bars, graphs, and AI interface markings; her face may be softly obscured by glow for a mysterious look.
3. Wide stadium view: the dark-haired girl watches from the foreground while the silver-blue-haired girl passes the glowing ball toward the lavender-pink-haired girl across a field covered in neon circuit lines.
4. Lavender-pink-haired girl flies forward low over the turf, chasing the ball through purple-blue stadium light and digital particle trails.
5. Close-up portrait of the lavender-pink-haired girl controlling the ball, long hair flowing dramatically, determined expression, surrounded by sparkles and holographic soccer energy.
6. Lavender-pink-haired girl rises in a burst of pink-purple magic, surrounded by many glowing heart particles and swirling petals.
7. Full-body shot of the lavender-pink-haired girl floating in midair inside a giant luminous heart-shaped aura, arms open, skirt and hair lifted by magical energy.
8. Frontal close-up of the lavender-pink-haired girl reaching both arms outward toward the viewer, with pink hearts and glittering particles behind her; her face may be covered by a soft purple placeholder glow for a stylized anonymous effect.
9. Reaction montage panel with exactly 4 internal sections: top section shows cheering anime spectators with surprised happy faces, bottom-left section shows two or three close-up spectator faces with soft face-obscuring glow, bottom-right section shows the humanoid AI robot mascot with glowing heart eyes, and pink heart icons fill the background.
10. Lavender-pink-haired girl performs an airborne kick or pass toward the floating star soccer ball, surrounded by glittering pink heart trails and a curved stadium horizon.
11. Dramatic long shot from behind three small opposing players on the field as an enormous neon pink heart-shaped goal effect explodes in the stadium, with the ball or energy point glowing at the center of the heart.
12. Final celebration: the 3 main anime soccer girls stand close together in the foreground with arms raised and smiling poses, framed by a glowing heart above them, flower petals, crowd lights, and purple-pink sparkles; faces can be softly stylized with glow or blur if desired.

Visual style: Highly detailed anime illustration, cinematic lighting, saturated neon sci-fi sports atmosphere, sparkling magical-girl effects, futuristic stadium packed with a cheering crowd, holographic data overlays, speed lines, petals, hearts, and particle effects. Make the image feel like still frames from a dynamic AI-generated anime soccer music video.

Text content: Only include the panel numbers 1–12; no captions, speech bubbles, logos, watermarks, or extra readable text.

Constraints: Preserve exactly 12 panels, exactly 3 main girls, exactly 1 soccer ball design, and exactly 1 robot mascot. Keep the scene wholesome, energetic, and cute. Use {argument name="primary color theme" default="violet, blue, and pink neon"}, {argument name="sport" default="soccer"}, {argument name="magic motif" default="glowing hearts and flower petals"}, and {argument name="setting" default="futuristic AI stadium"}.

Create a realistic vertical smartphone photo in a modern cinema lobby at night, showing an anime-inspired young woman named {argument name="character name" default="Emilia"} standing in three-quarter profile beneath a large digital “NOW SHOWING” schedule board. She has very long {argument name="hair color" default="platinum blonde"} hair in twin ponytails tied with black ribbons, pale skin, a slim figure, and her face is intentionally covered by a plain vertical beige rectangular privacy block. She poses thoughtfully with one hand raised near her cheek and the other arm folded across her waist, looking upward toward the movie listings. Her outfit is a gothic punk black ensemble with exactly 6 main visible pieces: an off-shoulder cropped black graphic T-shirt with white skull lettering, a black choker, a high-waisted black pleated mini skirt, a black grommet belt, two silver chain drapes at the hip, and black thigh garter straps. The cinema background should include exactly 5 Japanese movie titles on the digital board with showtimes 10:15, 12:30, 12:45, 14:45, 17:00, and 19:30, plus exactly 2 large illuminated upper poster panels labeled 3 and 4, exactly 2 framed lower movie posters on the wall, and red rope queue barriers with black stanchions. Use warm indoor theater lighting, glossy dark marble floors with reflections, shallow depth of field, slight handheld candid composition, realistic skin and hair texture, cinematic bokeh, and a photorealistic fashion snapshot style. Keep the scene in a {argument name="location" default="Japanese movie theater lobby"}, with Japanese signage such as {argument name="board heading" default="NOW SHOWING"} clearly visible but not overly crisp.

Create a vertical 2:3 fantasy anime-style illustration of one sorrowful young woman in left-facing side profile, standing under a star-filled midnight sky. She has very long flowing {argument name="hair color" default="deep blue-black"} hair streaming across the image, with countless fine strands, and her face is intentionally covered by one plain soft-edged taupe-gray rectangular blur centered over the facial area. She wears an ornate translucent layered {argument name="robe color" default="deep teal and icy blue"} robe with wide sleeves, white lace cuffs and collar, subtle gold floral embroidery, and a gauzy veil-like train blowing to the right. Add one elaborate silver filigree wing-shaped hair ornament behind her ear, decorated with multiple pearl beads, plus one dangling earring. In her hands she holds exactly one large pale pink peony blossom with green leaves and exactly two long pink ribbon tails hanging from the stem. The background is a dreamy cosmic night: dense blue nebula texture, many tiny snow-like specks, bright starbursts scattered throughout, and exactly one full moon in the upper right. Use delicate shoujo manga linework, luminous watercolor textures, intricate fabric and hair detail, melancholic romantic mood, cool blue palette with soft pink flower contrast, sparkling highlights, no text, no watermark.

Goal: Create a fictional weekly Japanese manga magazine cover in a bright summer shonen style, like a glossy newsstand issue, featuring an ocean-beach illustration as the main cover art and many energetic cover lines.

Canvas: Portrait magazine cover, approximately 2:3 aspect ratio, full bleed, saturated cyan-blue sky and sea, bold yellow-and-blue typography, high-density print layout.

Main cover art: A single anime-style teenage girl stands on a sunny tropical beach, seen from behind and slightly turned, holding a wide straw hat with a blue ribbon as wind blows her medium brown hair to the right. She wears a plain white short-sleeve T-shirt tucked into a high-waisted white summer skirt with pale blue floral patterns. Her face is obscured by a rectangular censor-like blur. The background shows sparkling turquoise ocean waves, white foam on sand, distant green island cliffs, a bright blue sky, puffy summer clouds, and palm leaves entering from the top right.

Magazine masthead and top area: Huge distressed italic masthead {argument name="magazine title" default="Mushin"} across the top in bright yellow with dark navy outline and rough halftone texture. Add a small weekly label at top left, a short Japanese-style tagline across the top, and a small release-day badge at top right. Issue information on the upper right should show {argument name="issue number" default="No. 28"}, {argument name="issue date" default="2025.7.2"}, and a price line of 550 yen.

Large main headline: On the right side, place a tall vertical Japanese summer headline reading {argument name="main cover headline" default="夏が呼んでる。"} in large white characters with blue shadow, plus a smaller pink vertical side phrase suggesting the dazzling smile of summer.

Inset feature boxes: Include exactly 3 inset feature panels. Panel 1 is on the left upper-middle: a dark blue action manga hero with spiky hair and a glowing blue fist, face blurred, framed in white, with bold blue Japanese feature text. Panel 2 is on the left middle: a pastel lavender-haired anime girl portrait with purple eyes, framed in white, with pink and blue Japanese feature text. Panel 3 is on the lower right: a faux gravure/interview promo showing a young woman in a white dress inside a glasshouse or garden setting, face blurred, with ornate serif title styling and floral decorations; include text implying an exclusive interview.

Other cover text: Add a left column with exactly 5 smaller serialized-title blurbs in stacked blue Japanese text, ending with a small pink “and more!!” style line. At the bottom, create a large campaign banner with a pink starburst badge on the left and the headline {argument name="bottom campaign text" default="人気キャラ総選挙2025 開催中!!"} in big blue, pink, and black lettering, followed by a smaller yellow-and-white voting call-to-action strip.

Visual style: High-resolution polished anime magazine art, energetic shonen cover design, bright summer palette, thick outlines, halftone print textures, glossy highlights, dramatic overlapping typography, sticker-like badges, drop shadows, and dense editorial layout. Keep the composition crowded but readable, with the central beach girl dominating the page.

Constraints: Use exactly 1 main character and exactly 3 inset panels; blur or obscure all visible faces in the main character and inset panels; do not use real manga titles, real publisher logos, or real celebrity names; make all Japanese text look professionally typeset and magazine-like.

Using the provided reference image, enhance and upscale it into an ultra-premium cinematic-quality photo. Preserve the subject's identity, pose, framing, and overall composition 100%, including the kitten behind the foreground surface and the plain blue background. Recover realistic micro-details: sharp eyes, natural fur texture, fine whiskers, detailed ears, subtle facial features, and cleaner depth. Remove blur, pixelation, noise, and compression artifacts while keeping the image photorealistic, high contrast, studio-quality, and natural. Target {argument name="image quality" default="8K ultra-premium cinematic"} with {argument name="detail emphasis" default="sharp eyes, pores/fur, and realistic hair"}; do not change the cat into a different animal, alter the pose, add objects, add text, or modify the composition.

Goal: Create a vertical anime manga page set inside a modern cinema lobby at night, featuring {argument name="character name" default="a blonde gothic punk girl"} deciding what movie to watch.

Canvas: Tall smartphone manga layout, approximately 9:16, full bleed, high-detail color illustration with black comic panel borders and slightly angled panel divisions.

Layout: Use exactly 4 panels. Panel 1 is a large dominant panel occupying the full right side and most of the height, showing the girl in side profile from thighs up, looking thoughtfully upward with one finger near her lips. Panel 2 is the upper-left close-up panel of the same girl, partly covered by a flat opaque tan censor rectangle over the face area. Panel 3 is a small middle-left rear-view panel showing her walking through the cinema lobby from behind. Panel 4 is the lower-left close-up panel showing her seated or leaning, looking back over her shoulder with a mildly uncertain expression.

Subject details: The character is a slim teenage/young-adult anime girl with very long {argument name="hair color" default="pale blonde"} twin ponytails tied with black ribbons, soft bangs, fair skin, reddish eyes, glossy highlights, and a delicate face. Her outfit is gothic punk: black off-shoulder cropped top with a white skull and rock-style lettering, black choker with metal ring, black wristband, pleated black mini skirt, belt with silver chains and rings, black thigh straps/garters, sheer dark stockings, and small metal accessories. Her mood is hesitant, thoughtful, and slightly embarrassed.

Environment: A sleek multiplex cinema lobby with warm overhead lights, reflective floors, poster displays, ticket counters, and a large digital showtime board in the background. The board should include Japanese UI text such as 「上映中」 and several fictional movie listings with times; include exactly 4 visible movie-title rows: 「コードギアス 復活のロゼ」, 「デューン 砂の惑星」, 「バッドボーイズ RIDE OR DIE」, and 「陰陽師 0」. Add showtimes such as 12:15, 12:20, 12:25, and 12:40 on the right side of the board.

Text content: Include exactly 3 vertical white speech bubbles with black Japanese text. Bubble 1 in the upper-left panel says {argument name="first speech bubble" default="どれにしようかな…"}. Bubble 2 in the middle-left panel says {argument name="second speech bubble" default="うーん…やっぱりあれかな"}. Bubble 3 in the lower-left panel says {argument name="third speech bubble" default="どうしたの？そんなにじっと見て…"}.

Visual style: Polished Japanese anime illustration, cinematic lighting, detailed hair strands, glossy skin highlights, high-contrast black clothing, deep shadows, warm bokeh lights, crisp comic inking, modern manga panel composition, dramatic perspective, rich detail, no watermark.

Constraints: Keep exactly 4 panels and exactly 3 speech bubbles. Do not add extra characters in the foreground. Keep all text legible where possible, but allow slight natural distortion on the distant showtime board.