GPT Image 2 prompts

Ultra-realistic full-body portrait of a {argument name="character description" default="confident young professional woman"} integrated directly into a {argument name="comic style" default="Japanese manga"} page layout. She appears inside a tall vertical hero panel on the left, spanning almost the full page height, and feels visually connected to the surrounding story panels. She wears a {argument name="outfit" default="sleek charcoal trench coat layered over a fitted black turtleneck and tailored wide-leg trousers, with leather ankle boots and a structured work bag"}. Her pose is stylish and confident, one hand lightly adjusting her coat collar, the other holding the bag, body slightly turned with one leg forward.\n\nOn the right side, create smaller black-and-white manga panels showing a day in her life in Tokyo: train station, office desk, creative meeting, coffee break, rainy street, and rooftop skyline. Make the woman feel like the main character of the page, as if the realistic figure exists inside the manga story, not outside it. Use strong ink lines, screentones, speech bubbles, motion lines, and cinematic panel storytelling. Blend the realistic subject naturally into the comic narrative while keeping her more detailed and visually dominant.

Photorealistic vertical 9:16 iPhone selfie photo of a {argument name="subject" default="fictional Japanese female idol"}.

beautiful adult Japanese woman,
{argument name="age" default="22-25"} years old,
soft oval face,
large sparkling dark brown eyes,
gentle slightly downturned eyes,
small natural nose,
soft pink lips,
healthy natural skin,
visible pores,
subtle skin texture,
tiny imperfections,
natural facial asymmetry,

long dark brown silky hair,
soft loose waves,
wispy see-through bangs,

wearing a {argument name="outfit" default="cozy cream-white knit sweater"},

resting her cheek on one hand,
gentle eye contact,
soft shy smile,
cute relaxed expression,

sitting inside a cozy café,
warm natural daylight,
wooden tables in background,
soft background blur,

close-up selfie composition,
camera slightly above eye level,
head and upper body filling frame,
arm partially visible,
slightly imperfect framing,

ordinary iPhone photography,
realistic smartphone HDR,
natural lighting,
realistic hair strands,
subtle image compression,
tiny focus imperfections,
social media snapshot,
friend-like atmosphere,

ultra realistic,
photorealistic,
beautiful but believable,
authentic Japanese idol daily life,

not studio photography,
not magazine photography,
not cinematic,
not CGI,
not AI-looking,

negative:
anime, illustration, 3d render, plastic skin,
fashion editorial, glamour lighting,
supermodel pose, watermark, logo, text

Create an artistic 16:9 CHARACTER IDENTITY BOARD.

SUBJECT: use the reference image. name: {argument name="character name" default="Ruby"}. make color correction.

Create a hyper-cinematic 16:9 dark fantasy samurai combat storyboard in 16 panels drawn as a rough kinetic manga sketch with raw sakuga energy, aggressive unfinished linework, explosive motion readability, and intense cinematic framing. A lone female ronin

Goal: Create a wide side-by-side comparison image showing two cinematic generations of the same mythic punk warrior, demonstrating an over-sharpened raw look on the left versus a softer refined film look on the right.

Canvas: Extra-wide horizontal frame, about 2.7:1 aspect ratio, split into exactly 2 equal panels with a thin vertical white divider line at the center. No captions, no labels, no watermark.

Subject: In both panels, show {argument name="character name" default="Nezha"} as a fierce young adult East Asian warrior with spiky black hair, dark red and black leather-and-metal armor, bare midriff, wrist wraps, belts, and long red silk ribbons swirling around him. His face should be obscured by a simple square blur/mosaic block, matching the source image. He grips a long spear or staff with both hands, thrusting it diagonally toward the viewer, emitting supernatural fire.

Setting: A dark ancient temple or shrine interior with carved wooden architecture, gold statues, candles, smoke, worn floorboards, and warm amber torchlight. Background details include exactly 1 large gold seated statue centered deep in the left panel and exactly 1 gold guardian-like statue in the rear right panel. The environment should feel like a mythic action scene in a Tokyo-inspired NEZHA fantasy series.

Left panel: Raw GPT-style output look. Tight action close-up from the waist up, warrior angled left-to-right, spear crossing the frame horizontally. Red magical flame blooms from the spear tip near the upper right. Make the image intentionally over-sharpened and hyper-detailed: glossy skin, oily highlights, high micro-contrast, razor-sharp edges, wet-looking armor, exaggerated sweat, crunchy texture, CGI-like red flame, dramatic but slightly artificial HDR finish.

Right panel: Refined final-film look of the same scene. Slightly wider composition, warrior crouched or lunging in a lower stance, spear thrusting from center toward the right foreground with orange fire streaming out. Use softer lens rendering, natural motion blur, warmer orange flame, less glossy skin, less plastic texture, more atmospheric haze, shallow depth of field, cinematic grain, painterly realism, and believable physical lighting. Keep the right panel noticeably less sharp and less oily than the left.

Visual style: Dark fantasy action cinema, high-budget live-action concept art, warm amber and crimson palette, volumetric smoke, temple dust, ember particles, dynamic ribbons. The two panels should clearly contrast two visual treatments of the same {argument name="scene concept" default="Nezha fire spear attack inside a temple"}: left is over-rendered and artificial, right is cinematic and natural.

Constraints: Include exactly 2 panels, exactly 1 center divider, exactly 1 warrior in each panel, exactly 1 spear in each panel, and exactly 1 visible flame attack in each panel. Avoid text, logos, UI elements, extra characters, and additional split-screen sections.

Goal: Create a moody literary poster about oppressive summer weather and a poet’s dampened spirit.

Canvas: Wide horizontal composition, approximately 16:9, divided into two main panels: a painted illustration on the left taking about 60% of the width, and a black text panel on the right taking about 40% of the width. Add a thin dark green border around the full image.

Left illustration: Paint in an expressive late-19th-century symbolist / post-impressionist style with heavy brush texture, muted ochre, olive, blue-gray, and storm-yellow colors. Show an interior study at night or late afternoon. A weary male poet sits at a wooden desk, dressed in a dark suit and white shirt, leaning his head into one hand while writing with a pen in the other. His face is covered by one flat opaque brown square. Behind and above him, depict a ghostly female muse or weather spirit rising from the humid air; her face is also covered by one flat opaque brown square. Around her head, show a storm-cloud shape containing a cracked brain-like form and exactly 3 jagged yellow lightning bolts. At the far left, place a tall window with exactly 4 large panes; outside is a dull cityscape and gloomy sky. Perched on the inner windowsill is exactly 1 black raven, beak open as if calling, one claw lifted toward the glass.

Desk objects: Include exactly 6 visible desk items: 1 stack of 3 closed books at the lower left, 1 spread sheet of handwritten paper in front of the poet, 1 separate loose paper, 1 black inkwell near the center front, 1 clear drinking glass at the lower right, and 1 pen held by the poet. Keep the objects painterly and slightly distorted.

Right panel: Use a pure black background. At the top center, place exactly 1 pale square QR-code-like graphic. Center the poem beneath it in pale green, hand-lettered, slightly irregular serif text. Poem text must read exactly:
{argument name="poem text" default="the weather\noppressively hot and humid\ndescends\nlike a heavy wet blanket\ndampening the poet’s spirit"}

Credits text: Below the poem, add smaller pale green hand-lettered credits, centered, reading:
AI image created by
{argument name="artist name" default="Michael H. Lester"}
in collaboration with
{argument name="model credit" default="GPT Image 2 Low on\nNightcafestudio"}
twitter: {argument name="twitter handle" default="lmhlester"}

Bottom links: At the bottom of the right panel, add two centered pale green URL lines:
{argument name="bottom links" default="https://creator.nightcafe.studio/?ru/mhlester\nhttps://4-michael-lester.pixels.com/"}

Style constraints: Make the image feel humid, oppressive, melancholic, and poetic. Use visible brushstrokes, dark outlines, and a handmade poster aesthetic. Keep the right-side typography readable but organic. Do not add extra characters, extra birds, extra lightning bolts, extra desk objects, or additional text.

Create a dark fantasy anime-style full-body portrait of a tree-human girl standing in a snowy winter forest. The subject is a tall feminine dryad/cyborg-like character with a human upper silhouette but her skin and limbs are transformed into cracked bark, twisted roots, and bare branches. She has extremely long flowing {argument name="hair color" default="black"} hair cascading almost to the ground, partially tangled with the branches, and a blurred/obscured face with a calm, eerie presence. Her arms are raised high and have become two large gnarled tree branches forming an arch above her head, with tiny red berries and patches of snow on the twigs. Her legs are bark-covered trunks ending in exposed root-like feet spreading across the snow. She wears the remnants of a torn black school-uniform-inspired outfit: a shredded black blazer, white shirt, short pleated skirt, thigh straps, and one tactical pouch strapped to her left thigh. Add exactly two distinct accessories/items: a rectangular ID badge pinned to her chest with the visible number {argument name="badge number" default="77"}, and a pale paper or envelope caught on the raised right-side branch. Above her head floats a small dark oval halo or sign with red lettering reading {argument name="halo text" default="JIS"}. The setting is a dense, cold, blue-gray forest at night or dusk, with dark tree trunks in the background, falling snow, snow-covered ground, scattered rocks, and sparse red berries on low branches. Use dramatic cinematic lighting, high detail, realistic bark texture, delicate snow accumulation, moody depth of field, elegant horror atmosphere, vertical composition, centered full-body framing, no extra characters, no gore, no watermark.

Goal: Create a square futuristic medical-status illustration for a cute star mascot, styled like a sci-fi HUD screen about a migraine episode.

Canvas: 1:1 square image, dark navy blue background with glowing cyan interface borders, subtle scanlines, neon UI accents, high contrast, clean digital illustration.

Main layout: A large central hospital-bed scene framed by a transparent holographic medical dashboard. At the top, display a big glowing English headline: “STATUS : MIGRAINE”. In the top-right corner add small text “MEDICAL LOG : HOSHI-CHAN” with a tiny plus sign. Under the headline, place two lines of Japanese message text: “昨日は片頭痛で寝込んでおりました。” and “久しぶりに大変きつうございました。” with the phrase “大変きつう” highlighted in cyan.

Subject details: In the center, show {argument name="character name" default="Hoshi-chan"} as a simple white five-point star mascot lying in bed under a blue plaid blanket. The star has tiny black dot eyes, a small straight mouth, and one blue teardrop of sweat on the right cheek, looking exhausted. Around the head, draw exactly 8 red-orange migraine pain symbols: 5 jagged lightning zigzags and 3 spiral squiggles. Behind the character are blue pillows and a dim bedroom setting. Include exactly 2 bedside objects: a dark bedside lamp on the left table and a small glass of water beside it. On the right side, include a tissue box with one white tissue sticking out.

HUD elements: Add a left vertical sidebar with exactly 5 icons from top to bottom: white star icon, horizontal menu lines, heartbeat line, bar chart, small dot. Add a compact right-side pain monitor card with a red warning triangle, red heartbeat waveform, label “PAIN LEVEL”, and a row of red segmented bars almost full.

Bottom information area: Use exactly 3 stacked glowing cyan status rows. Row 1 has a red warning triangle icon and Japanese text “状態異常：片頭痛 / 強制睡眠 / 食欲低下”, plus a small red heartbeat waveform on the right. Row 2 has a cyan capsule pill icon and Japanese text “服薬：片頭痛薬”, plus a circular cyan checkmark on the right. Row 3 has a cyan calendar icon and Japanese text “2026/05/30：痛みと睡眠で終わりました”, plus a tiny pixel-art white star mascot on the right.

Footer: Add small centered text “HOSHICHAN SYSTEM” above larger “REST MODE”. Include decorative cyan diagonal ticks, a small pixel heart icon in the lower left, and a crescent moon with a plus sign in the lower right.

Visual style: Cute but clinical cyber-medical dashboard, anime mascot simplicity, deep blue monochrome lighting with cyan neon UI and red-orange pain accents. Soft glow, crisp vector-like interface lines, legible typography, no photorealism.

Customization: The main diagnosis should remain {argument name="status text" default="MIGRAINE"}; the date line should use {argument name="date" default="2026/05/30"}; the two-line message should preserve {argument name="Japanese message" default="昨日は片頭痛で寝込んでおりました。\n久しぶりに大変きつうございました。"}.

Constraints: Keep all visible text sharp and readable, use exactly 3 bottom status rows and exactly 8 red-orange pain symbols, do not add extra characters, watermarks, logos, or unrelated UI panels.

Create a dramatic wide cinematic anime fantasy illustration of an elderly male wizard casting fire magic. The wizard has long flowing {argument name="hair color" default="white silver"} hair and a full beard, wearing a loose dark black robe with broad sleeves. His face is intentionally obscured by a simple centered square censor block in warm brown, covering the facial features while leaving the hair and beard visible. Compose him slightly left of center, torso angled toward the viewer, with one hand extended forward toward the right side and fingers spread as if controlling a massive spell; his other hand is raised lower on the left side in a claw-like gesture. On the right half of the image, show one enormous swirling {argument name="magic element" default="orange fire"} orb or portal, like a miniature sun made of rotating flames, bright yellow core, molten orange rings, and licking flame trails spiraling around it. Add a second smaller visible flame mass near the lower left edge. Use exactly two prominent fire masses: the huge orb on the right and the smaller flame on the lower left. The background is a dark smoky cavern or stormy void with charcoal gray and black brushy textures, illuminated by fiery orange rim light and scattered glowing embers. Style: high-quality painterly anime key visual, dynamic action pose, strong contrast, cinematic lighting, energetic brush strokes, detailed flame motion, dramatic fantasy atmosphere, widescreen 16:9 composition, no text, no watermark.

Goal: Create a high-density presentation overview image showing exactly 9 miniature slides from a Japanese business training deck about {argument name="course topic" default="Codex slide creation workshop"}. The slides should look like a polished AI-generated corporate consulting deck, arranged as a 3 by 3 grid on a white background with thin gaps between slides, each slide in 16:9 landscape format.

Canvas: Wide rectangular collage, approximately 2:1 aspect ratio, crisp vector-like slide thumbnails, white and navy color palette with small accents in blue, cyan, red, and gray. Use clean Japanese corporate typography, tiny but structured text blocks, charts, cards, icons, and slide numbers. Add a subtle professional workshop identity, like a small square “CX” logo and the label “CODEX SLIDE WORKSHOP” on multiple slides.

Layout: Include exactly 9 distinct slides:
1. Title slide: minimalist white slide with a navy “CX” logo at left, small spaced text “CODEX SLIDE WORKSHOP,” bold Japanese title 「誰でも出来る！Codex スライド作成講座」, date 「2026.05.31」, and a small subtitle about AI workflow / slide production.
2. Skills slide: title 「SKILLとは？目的を達成するための“技伝のタレ”」 with dense analytical layout, left vertical icon strip, numbered skill list, two small scatter/line charts, one donut chart, and multiple navy callout boxes.
3. Formula slide: title 「キレイ式AIスライド作成実践開始」 showing a process framework with icon cards, a comparison table, a small bar chart, and a dark navy bottom message banner.
4. Workflow slide: title 「Codexの使い方の簡単な流れ」 with a horizontal sequence of exactly 8 process cards using simple icons such as target, folder, cube, document, CX logo, chat bubble, circular arrows, and checkmark; include small tables, a donut chart, and a dark navy side memo box.
5. Goal slide: title 「本日のゴール」 with three numbered goal blocks on the left, a radar chart in the middle, progress bars on the right, and a bottom section containing a checklist table and navy callout box.
6. Simple creation slide: title 「Codexで、簡単に色々出来ちゃう！」 with a left column of three icon cards, a central bar chart, a donut chart at right, and a row of six small capability cards along the bottom.
7. Usage confusion slide: title 「けど、どう使えばいいか分からない」 with a large slanted funnel/pyramid chart on the left containing four stages and percentages, plus a donut chart, a vertical bar chart, a bottom table, and a dark navy quote card.
8. Efficiency slide: title 「本当はCodexをゴリゴリ使って、いろんなことを効率化したい」 with several horizontal bar charts, a line chart trending upward, four lower KPI cards with red negative percentages, and a navy recommendation box.
9. Explanation slide: title 「Codex使ってますか？」 with three explanatory bullet sections on the left, a central donut chart, a right-side bar chart, and a bottom row of KPI cards showing large figures such as -45%, -38%, 1日→90分, and +52%.

Visual style: Japanese management-consulting slide aesthetic, dense information design, crisp dark navy headers, thin gray grid lines, rounded cards, small monochrome icons, data visualization elements, and consistent spacing. Use realistic slide thumbnails rather than a single readable slide; the text may be tiny but should feel authentic and structured.

Text content: Preserve the main Japanese titles exactly where visible, especially {argument name="main title" default="誰でも出来る！Codex スライド作成講座"}, {argument name="workshop label" default="CODEX SLIDE WORKSHOP"}, and section titles containing {argument name="product name" default="Codex"}. Japanese body text can be small pseudo-realistic business copy, but keep headings, percentages, and chart labels consistent with the deck theme.

Constraints: Show exactly 9 slides, no people, no photographs, no watermark, no browser UI, no mockup device frame. The final image should look like a screenshot of an impressive complete slide deck generated for a workshop announcement.

Using REFERENCE_0 as the artistic base, transform the scene from a colorful fantasy market into a warm, detailed Japanese sushi restaurant interior while keeping the same ornate storybook-anime watercolor feeling, dense linework, glowing lights, and cozy golden atmosphere.

Goal: Recompose the two-girl scene into an intimate sushi-counter moment.

Main changes:
- Replace the market stalls and floating colored bubbles with a traditional wooden sushi bar filled with hanging lanterns, shelves of dishes, wall scrolls, patterned tiles, noren curtains, and chefs working behind the counter.
- Keep exactly 2 anime girls, but change their roles and styling:
  1. Left girl: a {argument name="left character role" default="green-haired sushi waitress"} leaning forward from behind the counter and presenting food.
  2. Right girl: a {argument name="right character role" default="long pale-blue-haired customer"} seated at the counter in an elegant icy-blue kimono-style dress, watching the food.
- Change the cuisine to {argument name="cuisine type" default="sushi"}. Add a small wooden serving board with exactly 5 visible sushi pieces: red tuna nigiri, pale white fish nigiri, yellow egg nigiri, salmon nigiri, and one seaweed-wrapped gunkan piece with orange topping.
- Add tabletop details: one blue floral teacup, one small soy sauce dish, one black rectangular bento tray with exactly 3 compartments, and small illustrated menu cards.
- Add background staff and diners as soft supporting figures, not the focus.
- Apply {argument name="face treatment" default="flat opaque beige rectangular blocks covering the visible faces"} to the main girls and background people, as if the face areas are intentionally masked.

Style constraints: Preserve the highly detailed hand-drawn watercolor/colored-pencil texture, luminous warm lighting, decorative pattern density, soft fantasy-anime atmosphere, and slightly nostalgic DALL-E-3-like illustration quality. Avoid photorealism, avoid modern minimalism, and do not add extra main characters.

Create a vibrant high-detail anime illustration of two teenage fantasy academy students sitting on outdoor stone steps on a steep seaside town street under bright summer sunlight. The left character is a petite fox-eared girl with {argument name="left character hair color" default="long flowing white hair with vivid red underside streaks"}, large white animal ears with pink inner fur, a loose white hoodie with black drawstrings, a short black pleated skirt, red socks, and chunky black-and-white sneakers; she sits with knees bent, both hands raised in surprise. The right character is a slim boy with {argument name="right character hair color" default="messy short white hair with red bangs"}, wearing a white short-sleeve school shirt with a crest patch on the chest and sleeve, a loosened red plaid tie, black pants, a belt, a silver hip chain, black backpack straps, red socks, and matching chunky black-and-white sneakers; he sits facing the girl with both hands lifted as if startled. Add flat opaque pale pink rectangular censor blocks over both faces, exactly 2 blocks total. Include exactly 3 red-and-white manga exclamation marks between the characters to show surprise. Scatter exactly 8 loose handwritten paper sheets around the scene: 2 partially visible along the left edge, 1 floating at the upper right, 1 near the top-right edge, 1 on the lower center step, 1 near the right edge, and 2 on the lower-right steps, plus 1 open notebook on the lower-right step. Background: a cinematic blue ocean bay, distant coastline, town rooftops descending the hill, telephone poles and wires, leafy trees casting shade, a red sign on the left, and a building facade on the right. Use dynamic perspective, crisp cel-shaded anime rendering, saturated blue sky with fluffy clouds, strong rim light, detailed clothing folds, windblown hair, playful school-life fantasy mood, ultra-clean line art, 4:3 landscape composition, no watermark, no extra text.

Using REFERENCE_0 as the character and outfit base, generate a polished photorealistic upscale scene of the same receptionist seated at a front desk in a modern office or clinic reception area. Preserve the navy suit, white shirt, brown shoulder-length hair, gold name badge, and the anonymized/blurred face treatment. Change the pose from the character sheet into a seated working posture: upper body centered, one hand resting on paperwork on the desk and the other arm reaching slightly toward the side as if assisting a visitor. Add a warm, softly blurred reception background with light wood counters and neutral walls. Include exactly 6 visible desk/background additions: 1 clipboard with papers in front of her, 1 black pen on the papers, 1 open laptop at the right edge, 1 small black desk sign at the left foreground, 1 table lamp in the rear left, and 1 small plant or greenery near the lamp. Make it look like a natural professional photo, vertical 3:4 composition, shallow depth of field, soft indoor lighting, crisp fabric detail, realistic skin and hair rendering, no character sheet panels, no extra people, no watermark.

Create a vertical 9:16 ultra-detailed cinematic anime cosplay illustration of {argument name="character name" default="a magical steampunk idol avatar"} flying high above a futuristic city at night, seen from an extreme wide-angle selfie perspective as if she is reaching toward the viewer with one hand in the foreground. She wears an ornate purple and white magician costume inspired by a stage-show ringmaster: a very tall deep-purple top hat with a leather belt, metal buckle, gears, and brass ornaments; long flowing {argument name="hair color" default="silver-white"} hair blown dramatically by wind; purple detached sleeves with gold filigree trim; a frilly low-cut bodice with lace, ribbons, a large purple bow at the neck, star charm necklace, layered white ruffled skirt, garters, and glossy thigh-high stockings. Add exactly three glowing magical circle sigils: one large blue-white sigil near the bottom foreground, one purple-blue sigil to the right behind her, and one smaller floating portal-like sigil at the upper left, all filled with concentric rings, runes, pentagram geometry, and radiant neon light trails spiraling around her body. The background is a vast aerial cityscape full of blue and violet skyscraper lights, clouds far below, a starry sky and visible Milky Way, with a few meteor streaks. Use dramatic motion blur on the closest hand, dynamic foreshortening, luminous rim lighting, glossy fabric highlights, particles, sparks, and a high-fantasy sci-fi atmosphere. Place a flat opaque muted gray-purple square directly over the character’s face, centered and covering the entire face area, while leaving hair, hat, neck bow, and outfit visible. Color palette: {argument name="main color palette" default="deep purple, electric blue, white, silver, and gold"}. Render in polished gpt-image-2 anime-realism style, highly detailed, sharp subject, magical glow, no watermark, no readable text except abstract runes.

Create a realistic vertical portrait photo of {argument name="character description" default="a young East Asian woman with long straight black hair"} standing beside a tall window with sheer white curtains in a warm elegant hotel bedroom. She is shown from about mid-thigh upward in a three-quarter side profile facing left, softly backlit by morning sunlight through the curtains, holding a strand of hair near her cheek with one hand and resting the other hand near her waist. Her face should be intentionally obscured by a smooth rectangular blur/mosaic block for privacy. She wears a delicate {argument name="outfit style" default="white satin fantasy-inspired dress"} with pale blue ribbon accents: a halter-like blue-and-white collar with a small bow, fitted bodice, short asymmetrical skirt with a high side slit, detached sheer white sleeves tied with blue ribbons at the upper arms, and a long blue waist sash with tassel ornaments. Emphasize soft fabric texture, glossy satin, translucent sleeves, subtle lace details, and a gentle relaxed morning mood. The room background includes exactly five prominent furniture/decor elements: an upholstered wooden chair behind her, a round vanity table with an oval mirror, a bedside lamp, a white bed on the right, and a small round table in the lower right foreground. Use warm beige walls, classic hotel decor, shallow depth of field, natural window light, muted tones, photorealistic DSLR quality, 3:4 vertical composition. Avoid text, watermark, extra people, exaggerated anatomy, or cartoon styling.

Create a realistic vertical smartphone night street photo shot with an ultra-wide 3.8 mm focal length, close-up framing. The subject is {argument name="character description" default="a young androgynous person with short platinum-blond layered bob hair"} standing very close beside a brightly lit Japanese vending machine on the left side of the frame. The camera is held close at chest-to-face height with a wide-angle perspective, making the vending machine feel large and near while the dark city street recedes into blur on the right. The subject’s face is intentionally covered by a plain opaque square privacy block in muted beige-gray, centered over the face. They wear {argument name="outerwear" default="an oversized soft gray hoodie with long sleeves"} over {argument name="outfit details" default="a white iridescent, sparkly idol-style blouse and dress with rhinestone trim"}; one hand is raised near the vending machine buttons with the sleeve covering most of the hand. The vending machine contains visible rows of bottled drinks, especially Pocari Sweat bottles, glowing cold blue-white from internal lighting; show blue price labels including 130 yen, 140 yen, and 150 yen and a vertical row of round black selection buttons. Background: a narrow nighttime urban Japanese street with dark building silhouettes, scattered neon and shop lights, shallow depth of field, realistic reflections on the vending machine glass. Style: candid documentary phone snapshot, natural skin-and-fabric texture, cool vending-machine light on the subject, high ISO night realism, slight lens distortion from the ultra-wide lens, no beauty retouching, no watermark. Use a {argument name="aspect ratio" default="vertical 9:16"} composition.

Create a vertical anime-style watercolor portrait of {argument name="character name" default="Nikke"} standing outdoors beneath a vivid summer sky. Show a close-up three-quarter side view from the chest up, with the character facing left and gently holding a single pink-and-white silk tree blossom near her face. She has deep {argument name="hair color" default="indigo blue"} hair with violet highlights, long bangs, and a high ponytail flowing to the right; add a gold hair clip reading “AI” near the temple and multiple small ear piercings. Her face is mostly covered by a large centered opaque square overlay with a muted blue-to-dusty-pink vertical gradient, concealing the eyes, nose, and mouth. Dress her in a sleeveless high-neck {argument name="outfit color" default="teal blue"} top with soft painterly shadows. Surround her with exactly three prominent silk tree flower clusters: one small flower held in her left hand beside her cheek, one large blurred foreground cluster at the bottom right, and one flowering branch cluster in the upper right corner with green leaves. Use a slightly faded but powerful blue palette, cool white balance, humid summer atmosphere, increased green tones in the shadows, strong visible film grain, watercolor paper texture, expressive brush strokes, bright backlighting, soft cloud patches, and sparkling highlights. Composition should be intimate and poetic, portrait orientation 9:16, with no extra text except the “AI” hair clip.

Goal: Create a vertical Chinese restaurant promotional poster for {argument name="restaurant name" default="福祥南门涮肉"}, themed around Children’s Day, old Beijing hotpot, and Qing palace costume fun. Make it look like a polished luxury food-advertising poster with cinematic product photography and ornate Chinese festival design.

Canvas: Vertical 9:16 poster, rich dark crimson background with warm golden lighting, subtle bokeh, floating confetti, steam, and decorative floral accents near the bottom corners. Use a premium red-and-gold palace aesthetic.

Layout: Top left contains the restaurant logo and brand name, with the English subtitle “FUXIANG NANMEN OLD PEKING HOTPOT” beneath it, then the location text “秀水店”. The upper-left headline dominates the poster: a huge gold brush-calligraphy “61儿童节”. Beneath it place a red rounded decorative label with gold border reading “宫廷童趣 欢乐开涮”. The right half features one child dressed as a Qing dynasty palace princess, face intentionally covered by a smooth rectangular skin-tone privacy blur, wearing a vivid blue embroidered robe and large black imperial headdress decorated with turquoise-blue flowers, gold ornaments, beads, and dangling tassels. The child holds one red candied hawthorn skewer in the right hand and reaches toward the hotpot with the left hand.

Food and product scene: In the lower-left foreground place one ornate old Beijing copper hotpot with a tall chimney, red-and-gold cloisonné-style body, boiling broth, heavy steam, and one slice of red lamb being dipped by wooden chopsticks. In the lower-right foreground place one clear rectangular plastic tray filled with many neatly layered thin slices of red-and-white marbled lamb, surrounded by steam. Add one small red-and-gold product badge near the tray reading “苏尼特羔羊肉” and “价值39元”.

Bottom promotion area: Add one large ornate red rectangular banner with gold trim across the bottom. Text inside: “凭天安门预约门票或穿宫廷服饰到店用餐” on the first line, and a larger second line “每桌赠送价值39元苏尼特羔羊肉一份”. Below it add a centered gold plaque button with red text: {argument name="call to action" default="到店即享"}.

Visual style: Hyper-realistic commercial food photography combined with elegant Chinese poster typography, gold foil texture, dramatic warm rim light, shallow depth of field, thick steam, appetizing glossy lamb texture, ornate traditional patterns, festive luxury atmosphere. Keep all visible text crisp and correctly spelled in Chinese where specified.

Constraints: Include exactly these major counted elements: 1 child model, 1 imperial headdress, 1 candied hawthorn skewer, 1 copper hotpot, 1 pair of chopsticks, 1 lamb slice being dipped, 1 clear lamb tray, 1 small product badge, 1 bottom promotional banner, and 1 gold call-to-action plaque. Do not add extra people, extra food trays, QR codes, watermarks, or unrelated English text.

Create a cinematic dark-fantasy royal triumphal procession titled {argument name="theme" default="Triumphal Return"}. A full-body frontal view of two regal riders advancing slowly down a wet cobblestone avenue through a medieval capital city. The left rider is a noble queen or female knight on a white horse, wearing polished silver plate armor, an ornate white cloak, gold filigree details, and braided blonde hair; her face is deliberately smooth and featureless with a soft skin-toned blur. The right rider is a king or prince on a black horse, wearing an immaculate white ceremonial military coat with gold embroidery, medals, braided cords, high collar, white cape, dark curly shoulder-length hair, and a similarly featureless blurred face. The two horses are large and realistic, walking side by side, with elaborate dark leather tack, gold medallions, decorative breastplates, reins, and royal harness ornaments. On both sides of the street, show exactly six prominent armored guards visible in the foreground and midground, three on the left and three on the right, wearing dark steel armor and helmets, some holding spears or standards; their faces are also blurred or obscured. Behind them, dense cheering crowds raise arms and banners. Tall black-and-gold heraldic banners hang from stone buildings on both sides. In the far background, an enormous Gothic palace or cathedral complex rises with many towers, spires, arched windows, buttresses, and ornate stonework. Confetti and flower petals fill the air and scatter across the wet reflective street. Lighting is dramatic late-afternoon sunlight breaking through clouds, with backlit particles, warm highlights on armor, deep shadows, volumetric haze, and a grand epic atmosphere. Use ultra-realistic detail, cinematic composition, symmetrical central framing, low eye-level perspective, 4:5 vertical aspect ratio, shallow depth of field with the riders sharp and palace slightly softened. Color palette: ivory white, black, antique gold, dark stone gray, muted rose petals, and warm sunlight. Avoid modern objects, readable text, logos, or extra main riders. The mood should feel like a victorious royal homecoming after war, majestic and solemn.

Create a vibrant anime-style full-body illustration of {argument name="character name" default="a cheerful young schoolgirl"} standing on a sunlit paved sidewalk in a suburban park or roadside setting. She is viewed from a high, slightly wide-angle perspective with a playful lens-like distortion, looking up toward the camera with closed smiling eyes, rosy cheeks, and a huge delighted grin; her hands are clasped together near her mouth in an excited, bashful pose. She has {argument name="hair color and style" default="short blonde hair with soft bangs and twin side braids that fade into pink tips"}, with small pale blue bow ties at the braid ends and a white flower hair clip on the left side. Her outfit is casual and cute: a loose black T-shirt with partial bold white lettering, bright pink shorts, white crew socks, pink sneakers with white laces and toe caps, and a small pink backpack with teal straps visible behind her shoulders. The environment shows a stone-tile walkway curving into the distance, a road or path with guardrails on the left, a metal fence on the right, green trees and bushes, distant blue hills, and a bright blue sky with fluffy clouds. Use warm golden-hour sunlight streaming diagonally from the upper right, strong rim light, long shadows across the pavement, dappled leaf shadows, and cinematic orange sun rays. Style should be polished Japanese anime key visual, crisp clean linework, glossy cel shading, luminous highlights, saturated colors, expressive face, dynamic composition, highly detailed background, portrait orientation 2:3. Avoid text overlays, watermarks, extra characters, distorted hands, or cropped feet.

Create a polished anime-style vertical portrait of a young woman in an elegant Victorian-inspired maid outfit sitting indoors beside a fireplace, viewed from a high three-quarter angle as if the camera is looking down at her. The character has {argument name="hair color" default="bright golden-orange"} long flowing hair swept to one side with glossy strands, a braided silver-white headband across the crown, and floral hair ornaments with two red camellia-like flowers, one small white flower, green leaves, red ribbon bows, gold beads, and a long red tassel near her right ear. She has large luminous {argument name="eye color" default="pink-purple"} eyes, soft blush, parted glossy lips, and a wistful upward gaze toward the viewer, with one hand gently touching her cheek. Her outfit is a {argument name="outfit style" default="white blouse and deep red Victorian maid skirt"}: a white low-neck ruffled blouse with puffed sleeves, lace cuffs, subtle gold embroidery, a red embroidered choker with a small jewel, gold teardrop earrings, a delicate necklace, a beaded bracelet, and a deep red high-waisted skirt/apron with gold floral embroidery and white lace trim. The setting is a warm dark wooden room with exactly 5 visible background decor elements: a black fireplace grate with glowing orange fire, a mantel shelf, framed pictures on the mantel, brass or copper kettles and bowls on shelves, and a bright window on the right casting strong golden sunlight. Use dramatic warm backlighting from the window, high-contrast patches of sunlight and shadow across her face and clothing, tiny dust motes in the air, rich brown wood tones, soft cinematic depth of field, detailed fabric folds, glossy anime rendering, romantic Victorian atmosphere, no text, no watermark.

Create a loose fashion-illustration style drawing of {argument name="character name" default="a young Korean man"} sitting cross-legged on a plain white background, centered on the canvas. He has messy short {argument name="hair color" default="black"} hair and a deliberately obscured face: place a soft beige rectangular blur over the entire facial area so no facial features are visible. Dress him in an oversized, angular {argument name="outer garment color" default="vivid cobalt blue"} sleeveless hanbok-inspired vest or modern ceremonial jacket with broad squared shoulders, layered over a dark navy long-sleeve shirt, with loose black wide-leg trousers and dark navy slip-on shoes with white soles. His posture is relaxed and slightly hunched forward; one hand rests near his lap while the other hand is raised beside his chest holding {argument name="food item" default="a small piece of seaweed rice roll with white rice visible"}. Use expressive black ink contour lines, visible sketch construction strokes, marker-like flat color fills, watercolor texture, and high-fashion editorial croquis proportions. Keep the background empty off-white, no text, no props, no shadows beyond minimal grounding, and emphasize the bold blue garment, long limbs, and casual melancholy mood.

Create a polished anime-style vertical illustration of {argument name="character name" default="a young Victorian maid"} standing in a luxurious late-19th-century manor tea room at golden hour. The character is shown from the hips up to full body, turned three-quarters away from the viewer with her back and side profile visible, elegant and slightly coy posture, one hand raised delicately near her chest and the other resting by a teacup on a wooden tea table. She wears a classic black Victorian maid dress with long sleeves, glossy fabric folds, white cuffs, white collar, white ruffled apron trim running down the front and shoulders, a large white bow tied at the back of the waist, and a white maid headband. Give her {argument name="hair color" default="dark brown"} hair in a long low ponytail with braided side sections, tied with a white ribbon, flowing strands catching the sunlight. Obscure the face with a smooth square censor block centered over the face, softly blurred in warm skin tones, with no facial features visible. The room is richly detailed: dark carved wood paneling, ornate wallpaper, crown molding, a glowing brass chandelier with candle bulbs and crystal drops above, a large tall window on the left with cream draped curtains and tassels, sunset light streaming in, a wooden tea table holding a tiered dessert stand with strawberry cakes, a white teapot, a teacup and saucer, and small pastries. In the background include a marble fireplace with a warm fire, mantel decor, candles, a vintage clock, framed landscape paintings in ornate gold frames, a vase of pink roses, and a red upholstered chair with patterned pillows. Use warm amber lighting, soft rim light on the maid, refined romantic atmosphere, high-detail anime rendering, clean linework, painterly background, vertical composition, no visible text, no watermark, no extra characters.

Create a wide single-panel editorial illustration on an off-white textured paper background. Show an adult Korean male baseball player seated on the floor near the right side of the canvas, cross-legged, slightly hunched, eating a triangular rice ball held in one hand. His face must be intentionally anonymized with a vertical soft-edged beige blur rectangle covering the facial features, while the rest of the head shows short dark hair and ears. He wears a vivid royal-blue professional baseball jersey with dark blue undersleeves, black pants, blue sneakers, and a wristwatch; include a large white team wordmark across the chest resembling “doyo,” the jersey number “1” on the lower front, a yellow rectangular sponsor patch on the upper chest, Korean sponsor lettering on the left chest area, and a large white number on the sleeve. Place exactly 2 small objects on the floor to his left: one upright clear plastic water bottle with a blue label and blue cap, and one small empty white paper cup. Use loose ink outlines, watercolor and marker-style shading, subtle gray cast shadows, and a melancholic sports-cartoon mood. Keep the composition minimal with a lot of blank space on the left, a thin floor line behind the objects and figure, no extra people, no extra props, and no readable text beyond jersey-like markings. The athlete’s jersey color can be customized as {argument name="jersey color" default="royal blue"}, the jersey number as {argument name="jersey number" default="1"}, the anonymizing face cover as {argument name="face anonymization style" default="soft-edged beige blur rectangle"}, the food item as {argument name="food item" default="triangular rice ball"}, and the mood as {argument name="mood" default="melancholic and lonely"}.