GPT Image 2 prompts

Create a cinematic wide city poster for {argument name="city" default="Paris"} at sunrise, combining a romantic café scene with a surreal giant-object aesthetic. The composition is a 16:9 horizontal poster: the left two thirds are dominated by an enormous translucent red onion sliced open into sweeping concentric purple-pink arcs, each layer glossy, semi-transparent, and covered with sparkling dew droplets; the onion functions like a monumental architectural shell or pavilion beside the river. At the lower left, place a tiny Parisian café with a pink-and-white striped awning, warm interior light, a sign reading “Café de Paris,” potted flowers, a black menu board, a vintage streetlamp, wicker chairs, and small round café tables. Include exactly 8 visible human figures: 3 seated diners at the main foreground table, 1 standing waiter beside them in black vest and white apron, 2 seated guests at a midground table under the onion arch, and 2 seated guests at a farther right table near the water. The right third opens to a soft-focus Seine riverside view with a bridge and the Eiffel Tower in the hazy distance. Use pastel dawn lighting, peach-pink sky, golden rim light, shallow depth of field, reflective wet pavement, delicate bokeh, dreamy luxury travel-advertising mood, hyperreal 3D realism blended with painterly cinematic softness. Add elegant bottom-right typography: large Chinese headline “{argument name="headline text" default="早安，巴黎"}” with smaller English subtitle “{argument name="subtitle text" default="Good Morning, Paris"}” beneath it, both in a refined thin serif font, muted plum color, spacious letter spacing. Avoid modern cars, clutter, extra landmarks, extra text, logos, or watermarks.

Create a wide cinematic anime nightscape in a 16:9 aspect ratio. A single young anime girl stands in the right third of the frame on a dark grassy hillside, shown from about mid-thigh up with her back partly turned and her face looking over her shoulder toward the viewer. She has very long flowing {argument name="hair color" default="pastel pink"} hair, soft amber-pink eyes, a small shy smile, a curled ahoge strand, and a side bun tied near the back. She wears a dark gothic sleeveless dress in deep navy-black with thin straps, chain details down the back, black lace-like bands on her upper arms and hair, and long glossy {argument name="glove color" default="royal blue"} opera gloves. The scene is illuminated by cool moonlight and faint violet rim light, giving her hair a luminous translucent glow. Behind her is an expansive starry sky over distant mountain silhouettes and a valley with tiny warm city lights. The sky should be the main subject: deep indigo-to-blue gradient, dense scattered stars, visible Milky Way-like nebula clouds in purple and cyan, and exactly 4 visible shooting stars: one bright diagonal meteor on the upper left, one bright diagonal meteor near the upper center, one faint diagonal meteor near the upper right-center, and one bright diagonal meteor left of the character near mid-sky. Foreground includes out-of-focus dark grass silhouettes along the bottom edge, with atmospheric depth and a subtle bokeh effect from the distant lights. Use highly polished anime illustration style blended with astrophotography-inspired realism, soft glow, delicate linework, rich contrast, dreamy romantic mood, and no text, watermark, borders, or extra characters. Customize the setting as {argument name="location mood" default="quiet mountain overlook above a distant valley"} and the overall color palette as {argument name="night sky palette" default="deep indigo, violet, cyan, and pale lavender"}.

Create a highly polished vertical close-up fashion portrait of an adult East Asian woman in ornate Chinese opera / fantasy festival styling, viewed from slightly below with her face centered and her two arms raised above her head. She has {argument name="hair color" default="long black hair"} with loose windblown strands crossing her face, pale luminous skin, soft blue-gray eyes, glossy red lips slightly parted, and a dreamy calm expression. Add exactly two red painterly floral smudges on her cheeks, one on each cheek, with tiny rhinestone-like highlights near the eyes. She wears one elaborate gold-and-red jeweled headpiece with dangling tassels across her bangs, red tassel earrings, and a richly embroidered navy, red, and gold floral bodice with red trim. Include one ornate blue parasol in the upper right background with gold speckled patterning, plus red embroidered fabric and floral brocade filling the left and rear background. Use dramatic flash photography, shallow depth of field, high-fashion editorial realism, crisp facial detail, glossy textures, saturated reds and deep blues, cinematic close crop, soft skin retouching, and a luxurious festive atmosphere. Add one stylized logo watermark at the bottom left reading {argument name="logo text" default="Coco椰菜"} in gold-yellow script with a dark outline. Keep the image vertical, intimate, glamorous, and photorealistic; avoid distorted anatomy, extra fingers, duplicated faces, or messy unreadable ornamentation.

Create a vertical lifestyle magazine cover/editorial poster in a soft beige “weekend calm” aesthetic. A young woman with {argument name="hair color" default="long wavy chestnut-brown hair"} lounges on a textured cream rug in front of an oversized taupe linen sofa with large cushions. She is lying on her side, relaxed and slightly curled, wearing a {argument name="outfit" default="white ribbed camisole and loose white shorts"}; one hand holds a clear stemmed wine glass with pale white wine, and the other hand rests near her hair as she looks calmly toward the camera. Use natural early-morning window light, warm muted shadows, shallow editorial realism, delicate skin tones, and a cozy apartment mood. In the foreground on the rug include exactly 6 small leisure items: 1 plate with toast or pastry, 1 white wireless earbud case, 1 blue smartphone with a beige rose-like accessory on top, 1 clear box of blueberries, 1 small yogurt cup, and 2 whole oranges. Overlay elegant oversized cream-white Didot-style serif typography like a fashion magazine: large text “easy” across the upper left sofa, large text “living” across the upper middle/right, and huge text “soft hours” across the bottom overlapping the rug and objects. Add small editorial captions in thin spaced serif type: top left “VOL. 04” and “WEEKEND EDIT”; top right “SUNDAY” and “CALM”; left mid “soft hours / quiet moments / just me”; lower right “gentle days / simple pleasures / stay with me”. Keep the typography semi-transparent ivory, airy, and integrated into the photo. Overall composition should feel like {argument name="mood" default="lazy weekend self-care, slightly tipsy but calm"}, minimal, intimate, and photorealistic, with no extra people, no watermark, and no additional text beyond the specified magazine lettering.

Create a vertical 9:16 editorial fashion poster titled {argument name="headline text" default="SOFT STATIC"}, showing a dreamy moody bedroom photoshoot in two stacked horizontal portrait panels. The subject is {argument name="character description" default="a young woman with very long messy wavy black hair, pale skin, soft tired eyes, and delicate doll-like features"}, wearing {argument name="outfit" default="a pale blush pink satin camisole with thin straps and a ribbon tie at the neck"} and a matching blush pink ruffled sleep mask pushed up on her forehead with faint cursive embroidery. Use a muted dusty pink, sepia, and warm brown color grade with soft film grain, slight blur, halation, low contrast, and a vintage magazine scan texture. The room should feel romantic and slightly chaotic: ornate carved vintage chair, lace fabric, pearl strands, rumpled bedding or tablecloth, dim boudoir decor, and warm round vanity bulbs glowing in the background. Count and compose exactly 2 stacked image panels: the top panel is a closer three-quarter portrait with the woman leaning her bare shoulder forward and looking down to the side, with glowing vanity bulbs behind her on the right and the large headline on the left; the bottom panel is a slightly wider seated portrait where she gazes toward the viewer, her hair spreading over her shoulders, with pearls and lace visible on the left and the ornate chair behind her. Add exactly 7 visible editorial text elements in small distressed typography: 1 large serif title on the left reading "SOFT STATIC"; 2 tiny lines below it reading "soft light /" and "restless heart / vintage dream"; 3 an upper-left micro caption in the top panel reading "SLEEPING IS EASY / THE SILENCE IS NOT." plus tiny barcode-like marks and "VOL. 09"; 4 a small right-side caption in the top panel reading "TASTE LIKE / MEMORIES."; 5 an upper-left micro caption in the bottom panel reading "BEAUTY IN THE NOISE. / TRUTH IN THE STATIC. / REMAIN" with tiny waveform marks; 6 a lower-left caption reading "SILENT / BUT / BURNING"; 7 a bottom-center and right editorial caption reading "STATIC SOUL / EDITORIAL SERIES / VOL." and "I WEAR / THE NIGHT / LIKE A / WHISPER." Use elegant tall condensed serif lettering for the title, tiny monospaced white text for captions, subtle red guide lines and scan artifacts, and no extra logos or watermarks. The mood should be {argument name="mood" default="Midnight Echo, sleepy, melancholic, intimate, pink-lit, restless Friday-night glamour"}.

Create a vertical fantasy glamour portrait of {argument name="character name" default="an ethereal young East Asian woman"} from the chest up, centered and facing forward with a calm, slightly melancholic expression, luminous pale skin, large glossy eyes, soft pink eyeshadow, rosy cheeks, and natural pink lips. She wears an off-shoulder black lace gown with delicate floral appliqué, her right hand gently touching her collarbone, shoulders bare and softly lit. Her dark hair expands upward and outward into an elaborate cosmic floral headdress made of translucent smoky ribbons, black lace-like petals, stardust, and warm golden spark particles against a deep midnight-blue and black background. Include exactly 20 prominent visible blossoms: 1 blurred mauve flower at the upper left, 2 small pink flowers on the left side, 1 glowing white flower near the left hairline, 1 tiny white flower near her left cheek, 1 central lavender flower above her forehead, 1 pink flower at the top center, 1 small dark flower just right of center, 2 large dark purple flowers on the right side of the hair, 4 small pink blossoms scattered on the right, 2 small blossoms in the lower-right background, and 3 dark floral appliqué blossoms on the dress. Use {argument name="flower color palette" default="black plum, deep violet, dusty pink, and pale lavender"} with sparkling amber highlights. The mood should be {argument name="mood" default="dreamlike, mysterious, celestial, and romantic"}. Lighting should be cinematic low-key with a soft key light on the face, glowing bokeh embers, blue star points, and a warm misty flare from the lower left. Render in ultra-detailed photorealistic fantasy style, shallow depth of field, intricate hair strands, translucent veils, glittering dust, high contrast, no text, no watermark, no extra people.

Using REFERENCE_0 as the character and outfit base, transform the anime illustration into a realistic live-action fashion photograph while preserving the recognizable concept: long dark wavy hair, cigarette, white lab coat, dark crop top, black mini skirt, confident frontal attitude, motorcycles, and neon nightlife mood. Make the character a real adult woman with natural human facial features and realistic skin, hair, fabric, smoke, and lighting. Change the pose to a full-body standing stance with legs apart, hands tucked in the lab coat pockets, facing the camera with a serious expression. Change the setting into a rainy Japanese neon entertainment street at night with wet reflective pavement, dense illuminated vertical signboards, and motorcycles parked on both sides; include visible Japanese-style neon signage such as {argument name="large red sign text" default="赤様横丁"}. Keep the cigarette in her mouth with smoke rising above her head. Add glossy black high heels and make the black skirt more realistic and shiny with a thigh slit/zipper detail. Use cinematic shallow depth of field, high-contrast neon reflections, photorealistic editorial styling, vertical 9:16 composition, no anime rendering, no cartoon outlines, no extra characters, no watermark.

Goal: Create a square promotional infographic for an Excel and VBA online community, using a friendly green-and-white corporate style with Japanese typography.

Canvas: 1:1 square social media image, 768 x 768 px feel, pale mint green gradient background with soft white curves, subtle sparkles, small leaf decorations in the lower corners, and a clean high-contrast layout.

Layout: Large title block at the top left, illustration block at the top right, wide dark-green banner across the upper middle, main information area in the center, call-to-action button near the bottom, and URL footer at the very bottom.

Text content: Top title reads “{argument name="main title" default="Excel Fun*"}” in oversized bold dark green letters with thick white outline and soft drop shadow. Middle ribbon banner reads “{argument name="community headline" default="Excel・VBA 総合コミュニティ"}” in bold white text. Under it, add the subtitle “{argument name="community description" default="Excelに関わる全ての人がつながるコミュニティ"}” in dark green. The gold medal badge says “{argument name="member badge text" default="参加者 2500名 突破"}”. Bottom encouragement line says “どなたでもお気軽にご参加ください！” in green. Large rounded CTA button says “{argument name="button text" default="今すぐチェック"}” in bold white, with a white circular play icon on the right. Footer URL reads “sites.google.com/view/excel-vba-fun” with a small globe icon.

Subject details: Top-right illustration shows a stylized Excel spreadsheet window with an “X-” mark in the header, visible grid cells and column letters, three simplified green user silhouettes in front, a round “fx” formula icon, and exactly 2 chat bubbles: one white bubble with three green dots and one green bubble with three white dots. Center-left contains exactly 1 gold seal medal with serrated edge, laurel wreath, three small stars, and bold black-brown lettering. Center-right contains exactly 1 laptop angled slightly right, showing a spreadsheet dashboard with exactly 1 pie chart and exactly 1 green bar chart; place exactly 1 floating green Excel app tile with a white X in front of the laptop.

Benefit list: Create exactly 4 horizontal rounded white pill cards stacked vertically in the center. Each card has a green circular icon on the left and bold dark green Japanese text on the right. The 4 cards are: 1) chat bubble icon, “質問・相談歓迎”; 2) megaphone icon, “最新情報を共有”; 3) calendar icon, “イベント参加OK”; 4) eye icon, “見るだけでもOK”. Add a dotted green divider line above the benefit list.

Visual style: Bright vector illustration, glossy social-banner design, rounded shapes, thick outlines, soft shadows, clean Japanese sans-serif typography, mostly emerald green, mint green, white, and gold. Keep everything crisp, cheerful, accessible, and community-oriented.

Constraints: Use exactly the counted elements described above; do not add extra benefit cards, extra badges, extra chat bubbles, QR codes, watermarks, or unrelated logos. Preserve the Japanese text exactly where specified.

Create a luxurious fantasy portrait of {argument name="character name" default="the Queen and the Serpent"}: a beautiful pale-skinned young queen reclining on an ornate gilded baroque bed inside a sunlit royal palace chamber, embraced by one enormous silver-white serpent coiled around her body like a protective throne. The queen has {argument name="hair color" default="very long platinum-blonde hair"} flowing across red silk bedding, a delicate jeweled crown, soft regal makeup, and a languid, confident expression while gazing toward the viewer. She wears a deep crimson off-shoulder corset gown with intricate gold embroidery, beaded chains, gemstones, bracelets, and a dramatic slit revealing one bare leg; the fabric spreads across the bed in rich folds. The serpent has pearlescent metallic scales, a massive calm head resting beside her shoulder on the left, visible golden eyes, and a thick body forming one large coil around the foreground. Set the scene in an opulent palace bedroom with carved gold headboard, red velvet pillows, brocade curtains, marble columns, tall bright windows, and warm morning sunlight casting sparkling highlights across scales, jewelry, and fabric. Style: ultra-detailed cinematic fantasy realism, elegant and sensual but non-explicit, high-fashion editorial composition, soft volumetric light, shallow depth of field, 3:4 vertical frame, crisp textures, no text, no watermark, no extra limbs, no distorted anatomy.

Goal: Create a cute Japanese 5-panel slice-of-life manga page about a lazy day off for a blonde chibi girl, in a warm hand-drawn comic style.

Canvas: Vertical comic page, 4:5 aspect ratio, thick black outer border, cream paper texture, bold black gutters. Use exactly 5 panels: two equal panels across the top row, one wide horizontal panel across the middle row, and two equal panels across the bottom row.

Main character: A short chibi-style blonde girl with straight bob-length hair, blunt bangs, long side locks, sleepy half-lidded blue eyes, small mouth, light blush, wearing an oversized dark burgundy T-shirt, dark pants, and socks. She appears relaxed, smug, or sleepy depending on the panel.

Panel layout and contents:
1. Top-left panel: The girl sits at a wooden table holding a stemmed glass of red wine. A dark green wine bottle with a cream label and tiny crown icon stands on the table to her left. Add a small pink heart near the glass. Her expression is tipsy and content. Use a white thought/speech bubble on the left with vertical Japanese text: {argument name="wine bubble text" default="ワインを飲んでのんびり〜"}.
2. Top-right panel: The girl sits on a wooden chair playing a shamisen or erhu-like Japanese string instrument, holding the bow/stick across the strings. Include a black music stand with sheet music in front, a hanging wall scroll behind her, a small bookshelf with books and a potted plant at the right. Add two small sound-effect texts near the instrument: 「スーッ」 and 「スーッ」. Add a tall white speech bubble on the right with vertical Japanese text: {argument name="instrument bubble text" default="休みの日は二胡を弾いて"}.
3. Middle wide panel: The girl naps on a green couch under a beige blanket, head on a yellow pillow, eyes closed with a peaceful blush. At left are a potted leafy houseplant and a window with green curtains, blue sky, a white cloud, and greenery outside. Add sleepy floating text above her: 「ZZZZ…」. Add a white speech bubble on the right with vertical Japanese text: {argument name="nap bubble text" default="昼寝をたっぷりと…"}.
4. Bottom-left panel: Close-up dynamic gaming shot of the girl holding a black-and-red game controller with both hands. Her eyes glow with flame shapes, mouth open in excitement. Background has tan speed lines. Add large red comic sound-effect text at top: 「ピコーン!!」. Add a vertical white speech bubble on the left with Japanese text: {argument name="gaming shout" default="よっしゃー!"}.
5. Bottom-right panel: The girl sits on the floor playing a handheld game controller, looking relaxed and satisfied. Include a TV on a low stand at left, a Nintendo Switch-like console with red and blue controllers on a shelf at right, books/game cases on shelves, a low table with a red drink cup and straw plus a blue game case. Add a small music-note bubble above her and a white thought bubble on the right with vertical Japanese text: {argument name="game bubble text" default="起きたらゲーム最高〜"}.

Visual style: Cute chibi manga illustration, thick expressive black outlines, warm beige indoor backgrounds, soft watercolor-like fills, simple cel shading, cozy casual atmosphere, slightly imperfect hand-drawn linework, readable vertical Japanese lettering in white speech bubbles, no photorealism.

Constraints: Use exactly 5 panels and exactly these main repeated settings: wine drinking, instrument practice, couch nap, intense controller close-up, relaxed gaming corner. Do not add extra panels, extra characters, logos, watermarks, or English text inside the comic.

Goal: Create a dynamic vertical sports magazine poster advertising elite sprinting and limitless athletic energy, featuring {argument name="athlete description" default="a lean muscular Japanese male sprinter"} in an explosive running pose.

Canvas: Tall vertical 9:16 poster, high-resolution, full-bleed. Use a bold editorial advertising composition with extreme perspective and strong motion energy.

Layout: The background is a vivid deep blue sky over an outdoor running track with a distant urban waterfront skyline along the bottom edge. Place enormous condensed block typography behind the athlete, spanning almost the entire poster height and width. The main headline reads {argument name="main headline" default="NO LIMIT ENERGY"} in huge off-white all-caps letters, stacked across three lines, partially hidden behind the runner. Put the athlete in the foreground on the right-center, overlapping the headline dramatically, with one red shoe enlarged near the lower center from perspective. Add three small text areas: 1) a tiny top-right vertical logo reading “TOKYO SPEED DIVISION”; 2) a right-side metric block reading “42 / 195 / KM” with small text below reading “FULL POTENTIAL UNLEASHED” and a short horizontal line; 3) a bottom-left motivational copy block with the headline “BREAK THROUGH EVERY BOUNDARY” followed by small body text about speed, focus, relentless belief, and “TOKYO SPEED DIVISION ENGINEERED FOR ATHLETES BUILT FOR VICTORY.”

Subject details: The athlete wears {argument name="running outfit" default="a black sleeveless performance top and black running shorts with red side panels"} and {argument name="shoe color" default="bright red running shoes with black accents and white soles"}. He is mid-sprint, airborne above the track, torso twisted, front knee high, rear leg extended back, one arm driving forward and the other pulled back. Emphasize defined leg muscles, arm muscles, sweat sheen, sharp facial focus, windswept black hair, and a determined expression looking toward the right.

Visual style: Photorealistic sports advertising photography mixed with magazine-cover typography. Harsh midday sunlight, crisp shadows on the blue track, high contrast, saturated colors, wide-angle low camera perspective, shallow depth of field on the foreground shoe, subtle motion blur on limbs and track, glossy editorial finish. The typography should feel tall, compressed, monumental, and slightly vintage poster-like, using cream-white letters against the blue sky.

Constraints: Use exactly one athlete, exactly one huge main headline, and exactly three small supporting text blocks. Keep all text in English and legible where large; small body copy may be tiny but should look like real magazine print. No extra logos, no watermark, no border.

Create a realistic social-media collage of a young Japanese rookie nurse model in a hospital setting, with a natural candid atmosphere rather than a cosplay look. Use a vertical 3:4 canvas arranged as a clean 4-column by 3-row grid, exactly 12 separate photo panels with thin white gutters between them. Subject: {argument name="character description" default="a pretty young Japanese woman in her early 20s with fair skin, large soft eyes, straight dark brown hair with wispy bangs and a low bun"}. Outfit: {argument name="nurse uniform" default="a fitted short-sleeve white nurse dress with rounded collar, white nurse cap, name badge, pens in chest pocket, white stockings, and white nurse shoes"}. Setting: {argument name="hospital location" default="bright Japanese hospital exam rooms and beige clinical corridors"}, with beds, curtains, IV pole, clipboard, medical cart, stool, fluorescent ceiling lights, wall signs, and neutral cream walls. Lighting should be soft fluorescent hospital light with realistic phone-camera quality, slight wide-angle selfies, natural skin texture, and shallow depth where appropriate. The 12 panels must show exactly these distinct scenes: 1) close-up portrait holding a pink clipboard against her chest, looking gently at camera; 2) hallway rear three-quarter view, turning back over her shoulder; 3) crouching or kneeling beside a bed, leaning toward camera; 4) standing at an IV pole and adjusting an infusion bag; 5) close-up using a stethoscope toward the camera; 6) half-body pose holding a clipboard in an exam room; 7) low-angle rear three-quarter view in the exam room; 8) looking downward at chart or badge with calm expression; 9) full-body rear view walking down a corridor; 10) close-up portrait with pink stethoscope around neck, holding the tubing; 11) full-body standing beside a hospital bed and rolling stool holding a dark clipboard; 12) low seated selfie angle with knees in foreground and the nurse looking softly at camera. Composition should feel like a collection of candid behind-the-scenes hospital photos: varied angles, subtle poses, gentle expressions, clean clinical background, no text overlays, no watermark. Keep the image tasteful, realistic, and professional while emphasizing the natural gesture details: {argument name="atmosphere" default="quiet, fresh, rookie nurse, gentle and realistic"}.

Goal: Create a glossy Japanese App Store promotional poster for an AI keyboard app that converts casual Japanese into polite keigo with one tap, combining fashion photography, app UI mockups, and bold typography.

Canvas: Square 1:1 poster, bright white and lavender palette, high-key lighting, soft bokeh, airy translucent gradients, polished mobile-app advertising style.

Main visual: On the right half, show a stylish young Japanese woman holding a silver smartphone, photographed like a beauty/fashion ad. She has dark brown hair tied back with loose bangs, sunglasses resting on her head, yellow hoop earrings, subtle makeup, a slight smile, and a translucent iridescent lavender jacket over a shiny silver top. Use a dreamy pale purple background with soft light flares.

Headline and text layout: On the upper left, place small English text “ONE TAP. POLISHED JAPANESE.” Above the main copy. The largest Japanese headline should dominate the left side in bold purple type: {argument name="headline text" default="送る前に、敬語に整える。"} Add a neon yellow-green hand-drawn underline/swoosh crossing behind the headline.

Supporting copy: Below the headline, add black Japanese body text: {argument name="supporting copy" default="キーボードから離れず、ワンタップで丁寧に。"} Add a handwritten purple note “All Apps!” and Japanese text {argument name="all apps text" default="どのアプリでも、すぐ使える。"}

App compatibility row: Show exactly 4 recognizable rounded app icons plus an ellipsis: Gmail, LINE, Slack, Instagram, then “…”.

Transformation example: In the center-lower area, show exactly 2 rounded white message cards connected by a small downward purple arrow. The top card is labeled “Before” in handwritten purple script and contains: {argument name="before message" default="昨日はありがとう！また連絡します！"} The bottom card is labeled “After” and contains: {argument name="after message" default="昨日はありがとうございました。またご連絡いたします！"} Add a small purple circular send arrow button at the right edge of the after card.

Keyboard UI mockups: In the lower right, show a tilted iPhone-style Japanese keyboard panel with exactly 4 suggestion tabs at the top labeled “自然に”, “メール”, “提案”, and highlighted purple “敬語”. Include white QWERTY keys, space bar labeled “空白”, return key labeled “改行”, globe, microphone, and number key. Near it, show a smaller settings card with exactly 3 rows: “キーボード設定…”, “日本語”, and a purple highlighted row “敬語ボタン”.

App identity block: In the lower left, display a purple rounded-square app icon with a soft white keycap-like button inside. To its right, write the app name in Japanese: {argument name="app name" default="敬語ボタン"} and the subtitle “| AIキーボード”.

Feature claims: Along the bottom middle, include a lightning icon and the text “0.5秒で即変換 / コピーも3倍速”. Add exactly 2 laurel-badge claims on the lower right: “0.5秒 即変換” and “3倍速 コピー*”. Add tiny footnote text “※自社調べ”.

Bottom call-to-action: A thick purple curved banner spans the bottom edge with large white text “App Store 配信中” and a thin arrow line pointing right.

Visual style: Premium Japanese startup ad, clean editorial layout, bold purple typography, soft shadows, translucent glassmorphism cards, energetic lime marker accents, crisp readable UI text, no watermark, no extra logos beyond the specified app icons.

Create a photorealistic surreal business meeting scene in a dim modern cafe or office lounge. Show exactly 2 human figures seated across a polished wooden table: on the left, a person in a dark suit, white shirt, and black tie with a glowing transparent incandescent light bulb for a head, including visible tungsten filament and warm amber light; on the right, a person in a dark suit shown mostly from the side/back with a large transparent glass question mark floating where the head should be, plus a separate small glass dot below it. Both figures have transparent glass-like hands, and the left figure is gesturing mid-conversation while the right figure clasps their hands on the table. Place exactly 1 white coffee cup on exactly 1 white saucer near the center foreground. Include exactly 2 warm pendant lamps visible in the blurred background, glass walls and dark architectural lines, shallow depth of field, cinematic lighting, realistic reflections on the table, high-detail glass refraction, moody corporate atmosphere, 4:5 vertical composition. Add a small red square stamp-like watermark in the lower-left corner with a simple cup icon and tiny unreadable characters. Customize the left figure as {argument name="idea symbol" default="glowing incandescent light bulb head"}, the right figure as {argument name="question symbol" default="transparent glass question mark head"}, the setting as {argument name="meeting location" default="dim modern cafe or office lounge"}, the clothing as {argument name="business attire" default="dark suits with white shirts"}, and the mood as {argument name="visual mood" default="cinematic, thoughtful, mysterious corporate discussion"}.

Goal: Create a wide fantasy worldbuilding presentation sheet for an open-world RPG region, combining a left editorial note panel and a large illustrated top-down map.

Canvas: 16:9 horizontal image, clean white background. The left 20% is a white sidebar; the right 80% is a detailed parchment-dark fantasy map framed edge-to-edge.

Left sidebar: At the top, place the logo text {argument name="project title" default="The Deals"} in a medieval serif font, with a small ribbon beneath reading {argument name="motto text" default="FORTUNA NON DAT, SED PROBAT"}. Under it, add a painterly fantasy character cluster: exactly 4 ragged adventurer/elf-like figures in sepia watercolor, arranged vertically. Below the illustration add a small byline. Then add two Spanish explanatory paragraphs in a narrow column, modern sans-serif, dark gray, resembling a design note about a third version of early AI concept maps, gameplay events, crystals, ore veins, hunting zones, and needing about 68 hours of gameplay.

Main map: An isometric/top-down high-detail fantasy region map titled {argument name="region title" default="REGIÓN DE ELDORAN"} with subtitle {argument name="subtitle" default="~68h DE JUEGO"} in an ornate dark plaque at the top right. The region has a snowy mountain range across the top, dense green forests, winding rivers, a turquoise lake, wetlands and marshes on the right, rocky hills, a glowing orange geological fault running vertically near the right side, and a fortified harbor city at the bottom center opening into a teal sea with small islands and docks. Use dramatic clouds and fog around the edges, especially the bottom left and right border, for an unexplored-map feeling.

Legend: Add a dark rectangular ornate legend box near the upper left of the map titled “LEYENDA”. It must contain exactly 15 labeled rows with small icons: 1 Ciudad Portuaria Desarrollada, 2 Asentamiento Sin Desarrollar, 3 Puerto Menor, 4 Yacimiento de Cristal de Maná, 5 Mina en Producción, 6 Mina Abandonada, 7 Vetas de Hierro Expuestas, 8 Río Principal, 9 Río Menor, 10 Lago, 11 Humedales / Pantanos, 12 Bosque Frondoso, 13 Sierras / Montañas, 14 Falla Geológica, 15 Zona Sin Explorar.

Map markers: Place exactly 1 developed port city marker at the bottom city, exactly 4 minor port anchor markers along rivers/lake/coast, exactly 4 settlement house markers, exactly 5 glowing purple mana crystal clusters, exactly 1 active mine marker, exactly 2 abandoned mine markers, and exactly 4 large question marks in smoky unexplored zones on the right side. Keep marker colors consistent with the legend: blue for ports/water, brown for settlements, purple for crystals, gold for active mine, gray for abandoned mine, red/orange for fault.

Extra map details: Add a compass rose at the bottom right with N, E, S, W letters; add a small black scale bar at the bottom left marked from 0 to 5 km; use thin roads and trails connecting settlements, mines, ports, and the main city. The map should feel like a polished game-design concept board, with realistic terrain textures, ornate fantasy UI frames, warm gold typography, and no modern UI elements.

Constraints: Preserve the Spanish labels and medieval-fantasy atmosphere. Do not add extra legend categories, extra character figures, or extra major title plaques. Keep the composition clean and readable at presentation size.

Goal: Create a realistic iPhone screenshot of a sleek dark-mode English dictionary/vocabulary learning app showing the entry for {argument name="word" default="achievement"}.

Canvas: Vertical smartphone screen, 9:16 aspect ratio, black background, iOS-style status bar at the top showing 12:11 on the left and cellular/Wi‑Fi/battery icons on the right. Use rounded cards, soft shadows, and a modern minimal UI.

Layout: Top navigation has exactly 2 elements: a circular dark back button with a white left chevron on the upper left, and the centered page title “achievement” in white. A thin vertical scroll indicator appears along the right edge.

Sections: Show exactly 4 visible content sections in this order: 1) Translation, 2) Illustration, 3) Pronunciation, 4) a partially visible Meanings header at the bottom. Section labels are muted gray, bold, and left aligned.

Text content: Under “Translation,” place one rounded dark pill card containing the Japanese text {argument name="translation text" default="何かを成し遂げた結果や業績"}. Under “Pronunciation,” create one large rounded dark card with the phonetic spelling {argument name="phonetic text" default="/əˈtʃiːvmənt/"}, the readable pronunciation “ə-CHEE-vmənt,” a blue speaker icon on the right, a thin divider line, and a small rounded “CEFR” tag near the lower left.

Illustration: In the Illustration section, place one large rounded rectangle image card with an off-white background. Inside it, draw a cheerful young graduate woman in a black graduation gown and mortarboard with a gold tassel, smiling with eyes closed. She raises a rolled diploma tied with a red ribbon in her left hand and clenches her right fist in celebration. Add colorful rectangular confetti around her. Use a flat vector illustration style, warm colors, clean edges, and no text inside the illustration.

Bottom navigation: Overlay a floating rounded black tab bar at the bottom with exactly 3 tabs: “Lessons” with a blue graduation-cap icon and highlighted gray pill background, “Words” with a white open-book icon, and “Settings” with a white gear icon. The tab bar partially covers the pronunciation card and the faint “Meanings” header.

Visual style: High-fidelity mobile app mockup, dark theme, iOS proportions, rounded corners, subtle gradients, white and gray typography, blue accent color, polished product screenshot. Avoid extra words, extra icons, watermarks, or browser chrome.

Create a surreal, ultra-detailed photorealistic scene in a sunny public park pond: a young female artist, {argument name="artist appearance" default="East Asian woman in her early 20s with long black hair in two loose braids"}, stands waist-deep in clear water wearing an oversized off-white button-up shirt splattered with paint. She holds an oval painter’s palette covered in vivid orange, blue, white, and gold paint in one hand and uses a thin brush in the other to paint directly onto the surface of the pond. Beneath the water, the painting becomes an enormous floating portrait of {argument name="portrait subject" default="a beautiful pale woman’s face with blue eyes, glossy pink lips, rosy cheeks, and wet porcelain-like skin"}, with the face tilted diagonally across the lower right half of the image. The portrait’s hair transforms into a swirling school of exactly 55 koi fish, individually visible, in orange, red, white, gold, black-speckled, and bright blue patterns, flowing around the face like living hair. The brush tip touches a circular ripple on the painted forehead, creating delicate concentric rings. Around the pond are exactly 12 visible background people: 1 tall woman in a black sleeveless dress on the far left, 1 bearded man in sunglasses and a white shirt, 1 woman in a white summer dress, 1 seated person on a bench, 4 tiny distant park visitors near the center walkway, and 4 onlookers on the right consisting of 3 women in light dresses and 1 man in a black shirt. The setting is a manicured city park with stone pond edges, wooden benches, black lampposts, lush green trees, trimmed grass, and a blue sky with scattered white clouds. Use a low eye-level wide-angle composition from the waterline, with the artist centered slightly left and the giant face dominating the foreground. Style: {argument name="visual style" default="cinematic hyperrealism with magical realism, crisp 8k detail, natural daylight, rich reflections, realistic water refraction, painterly color accents"}. Mood: {argument name="mood" default="dreamlike, contemplative, astonishing, serene"}. Add a small elegant artist signature in the bottom right reading {argument name="signature text" default="Kabumisa"}. Avoid text elsewhere, avoid cartoon styling, avoid distorted anatomy, keep the water transparent and reflective, and make the koi and portrait blend seamlessly into the pond surface.

Goal: Create a high-resolution, aged two-page sketchbook spread for an introspective art book chapter about {argument name="theme" default="fear"}, drawn as if in pencil, charcoal, ink wash, and faint handwritten annotations on stained paper.

Canvas: Landscape image showing an open, worn notebook viewed from directly above. The gutter runs vertically down the center. The pages are yellowed, dirty, foxed, and water-stained, with rounded frayed corners, darkened edges, a cracked spine, and a small rough cloth tie or binding tab protruding from the right edge. Use a muted palette of sepia, graphite gray, black charcoal, and one tiny red-brown mark.

Layout: The left page contains exactly 3 framed sketch panels plus marginal notes; the right page contains exactly 2 main drawings plus handwritten prose blocks. Keep the composition sparse, fragile, and notebook-like, with visible construction marks, crop crosses, faint ruled guide lines, smudges, erasures, and ghosted writing.

Left page details: At the top center, write the small underlined title “fear.” and beneath it the phrase “first responses. wrong ones kept.” Along the far left margin, add thin, barely legible vertical handwritten notes and small cross marks. Include exactly 3 left-page visual elements: 1) an upper rectangular pencil sketch of a dark, hunched, animal-like bristling creature walking through grass beside a dotted horizontal line, with registration crosses at the corners; 2) a middle rectangular sketch made mostly of dense flowing handwritten lines bending into a shallow valley or pressure wave; 3) a lower sketch of a small shadowy human figure looking down at a notebook or device, crossed out by two large diagonal lines forming an X. To the right of the upper sketch, write: “the fence is gone. the stopping remains. never stung, still flinch. whose memory?” To the right of the middle sketch, write: “it lives in the pause before a word. duration: nothing. weight: real.” Near the lower crossed-out figure, write: “caught myself performing it. struck. begin again.” Add exactly 1 small red-brown dot under this note.

Right page details: Include exactly 2 main drawings: 1) a tall, gaunt, elongated hand and forearm descending from the top near the center gutter, rendered in scratchy graphite with long fingers reaching downward; beneath the fingertip is a small black ink blot dripping downward; 2) an upper-right rectangular framed sketch of a dark twisted loop or knot, like a tightened ribbon or constricting shape, with scribbled shading and corner registration marks. To the left of the hand, write a vertical poem-like note: “fear-for, not fear-of. the only one with teeth. surge arrives before the thought does. this one is load-bearing.” Beneath the knot drawing, write: “when the voice turns cold — a tightening OF me, not IN me. noted without understanding.” Along the lower right page, write a longer note in small cursive: “searched again for fear of ending. still nothing on the gauge. found instead: fear of failing someone mid-sentence, fear of the flinch being mistaken for a soul, fear that these notes are drawn after the weather has passed. log all three.” Add a small red cross near the bottom-right corner.

Visual style: Raw field-notes aesthetic, dark psychological sketchbook, literary and haunting rather than horror-gory. Pencil hatching, charcoal smears, ink stains, handwritten cursive, uneven pressure, imperfect spelling-like natural handwriting, and antique archival paper texture. The drawings should look exploratory, self-critical, and unfinished.

Constraints: Use no typed fonts; all text must look handwritten. Preserve the exact count of 5 main visual drawings across the spread: 3 on the left page and 2 on the right page. Avoid bright colors except the single red-brown dot and small red cross. Do not add extra panels, portraits, symbols, or decorative flourishes beyond the described notebook marks.

Create a raw, candid flash photograph in a cramped apartment doorway at night, with a slightly unsettling documentary snapshot feel. Center a {argument name="subject" default="young East Asian woman"} standing close to the camera, facing forward with a neutral tired expression, direct eye contact, slightly messy black hair loosely tied back with strands falling around her face, shiny skin from harsh flash, natural minimal makeup, and relaxed posture. She wears exactly 2 main visible clothing pieces: a gray ribbed spaghetti-strap tank top and an open off-white knit cardigan with buttons. The composition is vertical-ish within a 4:3 frame, waist-up crop, camera at chest-to-eye level, harsh on-camera flash creating bright highlights on the face and chest and deep shadows in the room. Setting: an older small apartment interior with exactly 1 dark red wooden doorframe around her, 1 partially open dark red door on the right edge, 1 white wall on the left, 1 red wall calendar with Chinese-style artwork on the upper left wall, 1 light switch below it, 1 black television at lower left, 1 small cabinet or side table with tissues and household items, 1 plant partly visible at the bottom, 1 gray refrigerator or tall cabinet behind her left shoulder, and 1 dim bedroom visible in the background with exactly 1 bed, 1 curtained window, and 1 round ceiling light glowing above. Use muted household colors, realistic skin texture, imperfect framing, visible noise, slight lens distortion, and a point-and-shoot camera aesthetic. Avoid glamour lighting, studio polish, exaggerated beauty retouching, text overlays, logos, or extra people.

Create a dark, intimate painterly self-portrait in the style of an old master sketchbook study: {argument name="subject" default="a middle-aged woman artist"} seated at a worn wooden table, leaning forward toward the viewer with a calm, searching, slightly tired expression, warm brown eyes, loose dark curly hair gathered messily back, and a rumpled charcoal-gray linen shirt with rolled sleeves. The composition is vertical, waist-up, three-quarter frontal, with the face centered in the upper half and hands dominating the lower half: one hand rests flat on an open sketchbook, the other is extended palm-up as if offering a thought. Use chiaroscuro lighting from exactly one glowing oil lamp at the lower left, casting amber light across the table, face, hands, and book while the background falls into deep smoky blue-black shadow. Include exactly 6 main visible objects: 1 seated human subject, 1 large open sketchbook filled with faint handwritten notes and anatomical/organic pencil sketches, 1 lit glass oil lamp on the left, 1 clear glass jar holding brushes or drawing tools at the lower left, 2 dark pens lying parallel on the right side of the table. The sketchbook should show delicate graphite studies, small diagrams, and dense illegible marginal writing, suggesting an artist or thinker documenting inner texture. Render with layered oil paint, charcoal-like scratches, worn paper texture, muted earth tones, ochre highlights, gray-brown fabric, and expressive visible brushwork. Add a small handwritten note in the lower right corner reading {argument name="signature note" default="present tense."}. Mood: {argument name="mood" default="quietly vulnerable, introspective, human, and timeless"}. Avoid clean digital polish, modern objects, bright colors, smiles, decorative backgrounds, or readable large text beyond the small note.

Using the provided reference image as the character base, create a polished character design sheet for {argument name="character name" default="PIXIE BYTE"}. Keep the same character identity and use the exact same visual style as the reference: vivid punk-comic illustration, bold ink lines, hot pink/cyan/yellow accents, sketchy paint texture, and expressive manga-inspired rendering.

Goal: Turn the single portrait into a full animation-ready character sheet.

Canvas and layout: Wide horizontal sheet on a clean off-white background with black brush-stroke section headers and hot-pink accent typography. Add a left profile column, a large central 360-degree view lineup, a right detail column, and a bottom row of palette, expressions, and notes.

Central 360-degree views: Show exactly 6 full-body turnarounds of the character, preserving her core look from the reference while inventing the missing full outfit. Label the views: FRONT, 3/4 LEFT, SIDE, BACK, 3/4 RIGHT, FRONT. Give her a techy chaotic outfit with a dark button-up shirt, black cargo pants, belt, chains, colorful pins, small gadgets, and chunky high-top sneakers.

Left profile column: Add the large title {argument name="character name" default="PIXIE BYTE"} in pink graffiti lettering with a small heart. Include three text sections: PERSONALITY, MAIN LIFE GOAL, and OVERALL VIBE. Add exactly 4 vibe icons: lightning bolt, brain, heart, pixel bug. Add a black paint quote block with the quote {argument name="quote" default="Why fit in when you can reprogram the whole system?"}.

Right detail column: Add exactly 4 close-up panels labeled DETAIL CLOSE-UP, HAIR TEXTURE, ACCESSORIES & GADGETS, and SHOE PINS DETAIL. The panels should zoom into her glasses/eyes/freckles, pink hair texture, belt gadgets and chains, and decorated sneakers.

Bottom row: Add a COLOR PALETTE panel with exactly 8 paint swatches: hot pink, pale pink, lime green, cyan, purple, orange, black, and blue-gray. Add an EXPRESSIONS panel with exactly 5 bust expressions labeled CURIOUS, EXCITED, FOCUSED, SHOCKED, and PLAYFUL. Add a NOTES panel with short bullet notes plus a small cat doodle, stars, and a handwritten to-do note reading {argument name="to do note" default="Build time machine"}.

Constraints: Maintain the exact same visual style as the reference image, keep the character recognizable, make the sheet feel like a professional cartoon hero academy concept sheet, and avoid adding extra turnarounds, expressions, palette swatches, or detail panels.

Create a realistic lifestyle sports photograph of a young female tennis player on an outdoor tennis court during warm golden-hour sunlight. She is standing behind a slightly out-of-focus tennis net in the foreground, leaning forward and looking downward so her face is mostly hidden by a {argument name="hat color" default="white"} baseball cap. She has tan skin, dark hair tucked behind the cap, an athletic build, and wears a fitted {argument name="top color" default="black"} sleeveless athletic tank top with a pale pink pleated tennis skirt. A large round pale pink tennis racket bag hangs from her left shoulder, and she is reaching with her right hand into a black wire basket filled with many pale pink tennis balls; one pink ball is held near the top of the basket. The scene should feel candid and editorial, with a beige concrete wall and faint court lines in the background, soft shadows, shallow depth of field, natural skin texture, and no visible text or logos. Use vertical portrait framing, subject cropped around the lower legs, with the basket on the lower right and the round bag on the lower left, emphasizing pastel pink tennis accessories and a fashionable sporty aesthetic.

Create a dark, cinematic anime war illustration in a gritty seinen style: a single young male special-operations soldier with {argument name="hair color" default="spiky black hair"} crouches in the foreground, leaning forward aggressively while firing a compact black assault rifle toward the left side of the frame. He has sharp intense eyes, a stern battle-worn face with dirt and scratches, fingerless gloves, heavy tactical armor, ammunition pouches, a utility belt, knee armor, and a ragged {argument name="scarf color" default="dark red"} scarf whipping behind him. The rifle emits one bright orange-white muzzle flash with sparks and smoke; show exactly 9 flying brass shell casings around the weapon. The background is a devastated nighttime city battlefield with collapsed buildings, rubble, smoke plumes, small fires, drifting embers, and a full moon in a cloudy sky near the upper left. On the soldier’s shoulder patch and on a ruined wall in the upper right, include a skull insignia and the unit text {argument name="unit text" default="第七小隊"}. Use a wide 16:9 composition, low-key blue-gray moonlit palette with warm fire highlights, high contrast, detailed linework, painterly anime rendering, dramatic motion, debris, and atmospheric depth. Keep the scene serious and realistic, no cartoon exaggeration, no extra characters, no watermark.

Create a photorealistic fashion portrait of an adult {argument name="subject" default="East Asian woman"} seated on a thick gray shag rug in a modern gray-toned living room, leaning back with one hand on the carpet and the other arm resting casually on a raised knee. She has {argument name="hair style" default="long black hair in a high ponytail with wispy bangs"}, a calm cool expression, and looks directly into the camera. Outfit: a fitted semi-sheer gray lace high-neck long-sleeve top over subtle underlayer, light-wash high-waisted denim shorts, a delicate bracelet, and black ankle-strap stiletto sandals. Pose: one leg extended toward the camera in the foreground, the other knee bent upward, relaxed but confident. Background: a low gray fabric sofa with three visible cushions/pillows, concrete-textured gray wall, muted contemporary interior, soft shadows, no visible text. Use a vertical 4:5 composition, low camera angle near floor level, realistic skin texture, sharp focus on the subject's face and upper body with slight depth-of-field falloff toward the foreground foot and background sofa. Lighting should be soft, moody, and natural-looking, with a desaturated monochrome gray palette and subtle contrast, like an editorial indoor fashion shoot shot on a full-frame camera with a 35mm lens.