GPT Image 2 prompts

Create a cinematic wide promotional banner for {argument name="product name" default="GPT Image 2"} set in a dark enchanted forest at night. The scene shows one massive ancient tree trunk on the right with a hollow opening transformed into one glowing circular orange portal; through the portal is a misty blue mountain valley under stormy clouds. Surround the portal with warm sparks and amber rim light reflecting on twisted roots, wet stones, moss, and the forest floor. The left side should remain darker with tall tree silhouettes, blue fog, and negative space for typography. Add exactly 2 visible text elements in the lower-left corner: a large white headline reading “{argument name="headline text" default="GPT Image 2"}” and a smaller white subtitle below reading “{argument name="subtitle text" default="Now on Blyth AI"}”. Use a 16:9 landscape canvas, high-contrast teal-and-orange color grading, photorealistic fantasy environment, volumetric mist, shallow depth of field, dramatic cinematic lighting, crisp readable typography, no extra logos, no watermark, and no additional text.

Create a vertical 3:4 high-key creamy soft-haze studio portrait of a young woman in a strapless white draped dress, photographed from chest up against a warm beige seamless backdrop. She has {argument name="hair color" default="dark brown"} hair parted in the center and tied into a low, neat updo, with a small pearl earring visible. Her face is deliberately hidden by a centered opaque square privacy block in a soft beige skin-tone gradient, covering the entire facial area from forehead to chin. Use low contrast, overexposed milky highlights, feathered edges, and a translucent white veil or gauze foreground drifting along the right side and partially across the lower face and chest, as if shooting through white mist. Include exactly 2 visible pink orchid flowers: one larger orchid held delicately at the center of her chest by her hand, and one pink orchid tucked behind her left ear. Add a slim white orchid stem rising beside her neck. Her pose is gentle and editorial, shoulders bare, one hand holding the flower, with a delicate bracelet on the wrist. Lighting should feel like a white orchid soft-fog studio photoshoot: bright diffused key light from upper right, slight highlight bloom, creamy beige shadows, no harsh outlines. Add small white uppercase text {argument name="top logo text" default="NUYOAH VISION"} at the top left, and spaced white uppercase text {argument name="bottom caption text" default="SOFT HAZE STUDY"} centered near the bottom. Minimal fashion editorial composition, elegant, airy, romantic, realistic photography, no extra objects, no watermark beyond the specified text.

Create a high-end fashion magazine cover poster in a vertical 3:4 format. Use a minimalist studio fashion editorial aesthetic with a warm light-gray seamless backdrop and soft diffused lighting. The main subject is a young woman seated low on the floor in a relaxed, confident pose, photographed from a slightly low angle with dramatic perspective: one bent leg in oversized dark indigo jeans dominates the lower-left foreground, the other leg extends across the lower-right foreground, and her torso sits mid-right. She wears {argument name="top style" default="a white ribbed cropped tank top"}, very loose dark denim jeans, and translucent cream sunglasses resting on top of her head. Her hair is tousled medium-brown and wavy, with two visible colorful braided strands framing her face, using bright rainbow colors. Keep realistic skin texture and natural editorial retouching, not overly glossy or plastic. Her face is intentionally obscured by a centered soft-edged square privacy blur in warm skin-tone color, covering the facial features while leaving hair, sunglasses, body, and outfit visible. Compose the image like a niche fashion magazine cover with generous negative space in the upper-left and precise luxury typography. Add exactly 7 text elements: 1) top-left small uppercase text: "PRIVATE STUDIO EDIT"; 2) top-right small uppercase issue label: "ISSUE No.27" with "No.27" in muted red; 3) a huge elegant Didot/Bodoni-style serif cover title stacked on the upper-left reading {argument name="main cover title" default="OFF\nDUTY\nICON"}, with the letter O in "ICON" colored muted red while the rest is black; 4) mid-right small uppercase caption reading "MINIMAL\nMOOD." then a short muted-red horizontal dash, then "BOLD\nSIMPLICITY."; 5) left-lower small white uppercase caption beside a tiny muted-red dash reading "EFFORTLESS\nCONFIDENCE,\nNATURAL\nALLURE."; 6) bottom-center large serif slogan reading {argument name="bottom slogan" default="SIMPLE. STRONG. UNAPOLOGETIC."}, with only "UNAPOLOGETIC." in muted red and the rest white; 7) below the slogan, two centered lines of tiny spaced uppercase text reading "A PRIVATE EDIT FEATURING AUTHENTIC EXPRESSION\nAND MODERN FEMININITY." plus an even smaller centered footer reading "STUDIO ARCHIVE\n2024 COLLECTION". Add one vertical black-and-white barcode in the bottom-left corner. Use a refined black, white, and muted-red palette for typography, high-fashion magazine spacing, crisp print layout, shallow depth of field on the foreground denim, and no extra objects, logos, or text.

Create a vertical 2:3 mixed live-action and manga portrait showing an East Asian young woman taking a mirror-style selfie while her manga counterpart appears behind and around her. The real woman is on the lower right half of the frame in three-quarter side profile, holding a dark blue smartphone at the lower left edge with one hand visible. She has {argument name="hair color" default="black"} hair tied into a low messy bun with a large black claw clip, loose wispy bangs and flyaway strands, a visible ear, natural skin, and a calm introspective expression. She wears a fitted {argument name="outfit color" default="black"} V-neck cardigan or knit top with long sleeves. Behind her, fill the background with a huge black-and-white manga illustration of the same woman in matching side profile, matching bun, hair clip, long neck, V-neck top, and delicate linework, scaled so the manga head and hair dominate the upper and left areas. Add a comic contact-sheet border across the top and right side: exactly 7 small manga portrait panels total, each showing the same woman from slightly different side angles with the same hairstyle and outfit; arrange 5 rectangular panels in a horizontal row across the top and 2 smaller panels stacked vertically on the right. Use clean white background, thin black panel borders, detailed ink hatching, high-contrast monochrome manga lines, and subtle halftone texture. Blend the real subject and illustrated background so they align visually, as if the woman is taking a selfie in front of her own manga character sheet and the real and drawn versions share the same identity. Keep the composition bright and minimal, with no text, no watermark, no extra people, no exaggerated beauty retouching, and realistic East Asian facial proportions. Use {argument name="photo style" default="natural smartphone selfie realism"} for the live-action part and {argument name="manga style" default="delicate black-and-white shojo manga ink drawing"} for the illustrated part.

Using REFERENCE_0 as the kitten identity, create a clean WeChat-style sticker-pack preview sheet on a plain white background. Transform the real photo into a soft, semi-realistic cute 3D/cartoon gray kitten with plush fur, round proportions, and gentle expressive body language, while preserving the kitten’s overall gray color, small ears, baby face, and soft fragile look.

Layout: Arrange exactly 6 separate sticker poses in a 3-by-2 grid with generous white spacing and no borders.

Sticker poses, exactly 6:
1. Sitting upright facing forward, paws together, tail curled beside the body.
2. Crouching low and peeking forward with wide eyes, playful and timid.
3. Sitting upright with one paw raised high, showing pink-and-dark paw pads.
4. Curled up sleeping peacefully, eyes closed, tail wrapped around body.
5. Jumping upward with both paws forward, small motion lines and a faint oval shadow underneath.
6. Standing on hind legs with both paws raised outward, showing paw pads, cheerful greeting pose.

Add a solid medium-gray square placeholder over the face area of poses 1, 3, 5, and 6 only, as if censoring or reserving space for future text/expressions. Do not add any text. Keep the sheet polished, adorable, consistent in character design, and suitable for a finished sticker/emoticon set preview. Use {argument name="sticker count" default="6"} stickers, with {argument name="background color" default="plain white"}, in a {argument name="art style" default="soft semi-realistic cute 3D cartoon"} style, based on {argument name="character" default="the gray kitten from the reference photo"}.

Create a cinematic vertical photo of one energetic young adult female football fan/cheerleader in a packed stadium at night, shot from a low close-up angle from the seats as she turns sideways and leans forward with motion blur. She has long dark brown hair blowing in the wind, and her face is intentionally covered by a plain opaque gray square censor block. She wears a bright {argument name="outfit color" default="yellow"} sporty cheer outfit with exactly 2 visible clothing pieces: a short-sleeve cropped jersey top with dark green-and-white sleeve trim, and a matching pleated mini skirt with dark green-and-white side stripe accents. In her right hand she waves exactly 1 small {argument name="flag" default="Chinese national flag"} on a white pole, blurred by motion, with the red flag and yellow stars visible on the right side of the frame. The background is a huge football stadium filled with a dense crowd, glowing floodlights, misty haze, colorful bokeh, and a dramatic backlit halo around her hair. Use shallow depth of field, soft film grain, realistic skin and fabric texture, dynamic candid sports-event photography, high energy, celebratory atmosphere, slightly tilted composition, red stadium seats in the foreground, no readable text, no watermark.

Create a cinematic cold sci-fi scene in a vertical 4:5 frame from an extreme high-angle overhead camera, looking down into a vast oppressive concrete testing courtyard. The environment is a circular Chinese courtyard-inspired laboratory chamber made of wet dark stone, steel-gray concrete walls, mint-green cyan light accents, exposed black cables, drainage seams, and glossy rain-slick floor reflections. In the lower center foreground stands one small sharp figure: a cat-eared cyberpunk girl named {argument name="character name" default="VOXCAT"}, treated like an abnormal experimental sample rather than a cute mascot. She has pale mint hair, glowing translucent cat ears, a cropped white and mint tactical outfit, straps, thigh harnesses, arm sleeves, chunky boots, and a small shoulder marking reading “VOX CAT”; her face is deliberately obscured by a smooth gray anonymizing blur, making her feel monitored and dehumanized. The composition must emphasize environmental pressure: the character is small, centered below the huge circular arena, with massive walls and architectural rings dominating her. Include exactly one large circular test platform in the middle, one raised central disk, one dark sealed doorway at the far wall, two concrete blocks in the front left and front right foreground, one stylized cat/fox laboratory emblem on the upper left wall, one RX-07 emblem on the right foreground block, and one handwritten VOXCAT signature at the bottom right. Text elements: on the back wall place {argument name="main wall text" default="情感实验第7区"}, beneath it {argument name="wall subtitle" default="理性优先 情感禁止"}, and on the doorway lintel small text reading {argument name="door warning text" default="观测中｜非参与者禁止入内"}. On the central disk write “关系实验场 RX-07”. Around the circular floor ring place three visible Chinese warning labels: “接近＝风险”, “亲密＝变量”, and “情绪＝噪声”. Lighting should be low-key, teal-cyan and steel gray, with low saturation, hard concrete texture, wet specular highlights, volumetric haze, surveillance-camera unease, and a cinematic dystopian laboratory mood. Use a wide-angle lens, deep depth of field, high-position surveillance composition, no warm colors, no crowd, no extra characters, no cute cheerful atmosphere.

Goal: Create an elegant fictional Japanese art festival exhibition poster in a dreamy watercolor-and-photography mixed style, titled {argument name="main title" default="蒼花幻灯"}, with a poetic blue, cyan, coral, and ivory floral atmosphere.

Canvas: Vertical poster, 3:4 aspect ratio, high resolution, off-white paper background with subtle watercolor staining and fine grain. The composition should feel like a premium gallery/event poster, airy and romantic, with refined Japanese typography.

Layout: Place the large vertical Japanese main title on the left side, occupying the upper-left to mid-left area. Above it, add a small event line reading {argument name="event name" default="蒼花市芸術祭 2026"}. Beneath the title, add a short poetic subtitle in small Japanese text. Below that, place a thin decorative divider line, then the date range, venue, and location information. Put a tiny footer line along the bottom edge. On the right lower half, show one young woman in profile facing left, standing among flowers; her face is intentionally covered by a soft square beige blur/mask with no facial features visible.

Subject details: One female figure only. She has shoulder-length dark navy-black hair with vivid magenta-pink tips, slightly messy and backlit. She wears a translucent flowing kimono-like shawl or dress printed with blue and coral flowers. Her pose is calm and contemplative, head angled slightly upward, seen from the chest up. The figure blends naturally into the painted botanical environment.

Floral and watercolor design: Create a large diagonal cloud of watercolor blooms and ink washes sweeping from the upper left across the center to the right side, dominated by deep indigo, cobalt blue, turquoise, and cyan, with warm coral, peach, and orange blossoms scattered through it. Use many delicate thin botanical stems, tiny leaves, and small flowers overlaid like pressed flowers or ink illustrations. The bottom foreground is a soft field of wildflowers with blurred bokeh-like depth: coral cosmos-like flowers, blue blossoms, pale stems, and glowing highlights.

Text content: Include exactly 7 visible text blocks: 1) top small event name: {argument name="event name" default="蒼花市芸術祭 2026"}; 2) large main title: {argument name="main title" default="蒼花幻灯"}; 3) poetic subtitle: {argument name="subtitle" default="色彩が記憶に触れる、ひと夜限りの幻想展示"}; 4) date line: {argument name="date line" default="2026.10.17 Sat — 11.03 Tue"}; 5) venue line: "星雫ホール / HOSHIZUKU Hall"; 6) location line: "蒼花市・水庭区"; 7) bottom footer: "光、花、そして揺らぐ感情の輪郭を描く。    入場自由 / Exhibition Poster". Keep the Japanese typography crisp and legible, with the main title in a large elegant Mincho-style serif font and all smaller text in refined navy lettering.

Visual style: Luminous watercolor wash, alcohol ink textures, fine botanical line art, soft paper texture, cinematic shallow depth of field at the flower field, delicate museum-poster design, refined negative space. Use deep blue typography matching the indigo paint. The mood is quiet, magical, poetic, and exhibition-like.

Constraints: No logo, no QR code, no extra text beyond the 7 text blocks, no realistic sharp facial features because the face must remain covered by the square blur, no cluttered layout, maintain generous margins and a sophisticated Japanese poster aesthetic.

Goal: Create a vertical museum-gallery style Japanese explainer infographic about {argument name="headline text" default="Claude Fable 5に、米国が動いた"}, shown as a curated wall exhibit with ornate gold picture frames, warm spotlights, and a small viewer character observing the display.

Canvas: Portrait 3:4 composition, warm cream gallery wall, wooden floor, subtle shadows, two ceiling spotlights at the upper left and upper right, realistic 3D-rendered infographic style with elegant editorial lighting.

Layout: Large centered headline at the top in bold black serif Japanese typography, with small gold ornamental flourishes above and below. Beneath it, arrange exactly 3 main ornate framed exhibits connected left-to-right by exactly 2 gold arrows: one straight arrow from the left frame to the center frame, then an L-shaped arrow from the center frame down/right to the large right frame.

Main framed exhibits, exactly 3:
1. Left frame: small gold frame with a dark temple-like background and a carved stone tablet. Top navy label reads {argument name="left frame label" default="Mythos 5"}. Bottom ivory label reads 「高性能モデル」. Below this frame, attach one ornate ivory plaque containing a small brown treasure chest with a gold lock and the text 「フル性能だが限定提供」.
2. Center frame: medium gold frame showing an open illustrated storybook on a table, with a bright landscape illustration on the right page. Top navy label reads {argument name="center frame label" default="Fable 5"}. Bottom ivory label reads 「公開向け調整」. Below this frame, attach one ornate ivory plaque containing a simple cardboard box icon and the text 「危険部分を抑制」.
3. Right frame: largest gold frame on the right, deep red background, containing a sealed red envelope with a gold U.S. eagle seal. Top red label reads 「米国通達」, large central text reads 「停止通達」, and bottom red label reads 「全アクセス停止」. Below this frame, attach one ornate ivory plaque containing a red circular pause icon and the text 「国籍判定できず 全員停止」.

Additional text box: Place one small ornate ivory note card near the lower center-left, between the viewer and the right frame. It should contain exactly this Japanese body text: 「性能と安全のせめぎ合いが、国境の壁を生んだ。その判断が、すべての扉を閉じた。」 with a tiny gold ornament beneath.

Viewer character: Add exactly 1 chibi-style adult woman standing at the lower left, seen from behind in three-quarter view, looking up at the exhibit. She has {argument name="hair color" default="dark brown"} shoulder-length bobbed hair, glasses, a white blouse, black pencil skirt, and black flats. Keep her small relative to the wall display.

Visual style: Premium museum infographic, ornate gold frames, cream and burgundy palette, serif typography, soft realistic shadows, polished wood floor, balanced negative space, subtle depth-of-field, high-resolution vertical poster design.

Constraints: Use exactly 3 main framed exhibits, exactly 3 lower explanatory plaques attached to those exhibits, exactly 1 separate note card, exactly 2 directional arrows, exactly 1 viewer character, and exactly 2 ceiling spotlights. Preserve the Japanese text as written, avoid extra labels, avoid watermarks or logos, and keep all text legible.

Create a polished anime-style vertical illustration of a young woman enjoying a summer night at an open-air café terrace, themed around a “night sky soda” outfit. The character is shown from behind in a three-quarter seated pose, looking toward a deep blue starry sky and a distant city skyline across the water. She has long wavy dark navy hair tied in a high ponytail with a large pale blue bow decorated with tiny gold stars. Her face is mostly turned away and softly obscured, keeping the focus on the outfit and atmosphere. She wears an off-shoulder pale blue frilled blouse with delicate gold constellation embroidery, a loose dark navy cardigan slipped off her shoulders with shooting-star and star motifs on the back, and a layered pleated skirt in navy and light blue with scattered star embroidery. Add a matching frilled drawstring shoulder bag hanging from the chair, decorated with gold chains and exactly 5 visible charms: one miniature blue cream-soda glass charm, one gold star charm, one blue teardrop jewel, one round blue jewel, and one small white pearl-like charm. On the café table at left, place exactly 1 tall blue cream soda in a glass goblet with a straw, sparkling gradient blue liquid, ice cream scoop, whipped cream, and a red cherry, plus exactly 1 small candle lantern nearby. The background should include exactly 2 warm hanging lanterns on the right, leafy plants framing the terrace, a railing, blurred city lights, and a sky filled with many stars plus exactly 2 visible shooting stars. Use rich cinematic lighting, warm lantern highlights against cool moonlit blues, glossy reflections on the table and glass, intricate fabric detail, sparkling star accents, high-detail anime rendering, romantic summer-night mood, vertical 2:3 composition, no text, no watermark. Optional customization: make the main subject {argument name="character name" default="anonymous night-sky girl"}, use {argument name="hair color" default="dark navy"}, set the outfit theme to {argument name="outfit theme" default="night sky soda"}, set the drink to {argument name="featured drink" default="blue cream soda"}, and set the setting to {argument name="setting" default="open-air café terrace overlooking a city at night"}.

Create a vertical 9:16 high-fashion fantasy photo of {argument name="subject" default="a young woman"} crouching in a dreamlike floral environment, shot from a very low, wide-angle perspective as if the viewer is among the flowers. The subject has long flowing {argument name="hair color" default="warm brown hair"} blown outward by wind, wears a delicate pastel floral dress with sheer petal-like layers, and has small blossoms and twig-like floral ornaments woven into her hair. Her face is intentionally hidden by a flat vertical matte taupe rectangle centered over the face, while her right arm reaches dramatically toward the camera with an oversized foreshortened open hand in sharp foreground focus. Surround her with exactly six prominent large blossoms: one blurred peach flower in the lower left foreground, one bright cyan-blue flower near the lower center, one orange-yellow blurred flower at the lower right edge, one peach flower on the left middle background, one peach flower on the upper right background, and one small yellow-orange flower near the upper left. Add dozens of flowing translucent ribbons and paint-like tendrils in {argument name="color palette" default="coral pink, peach, aqua blue, turquoise, and soft orange"}, swirling around her hair and body like liquid petals. Use bright white backlighting, airy overexposure, soft bloom, shallow depth of field, pastel color grading, ethereal spring atmosphere, glossy cinematic detail, and a whimsical surreal editorial fashion aesthetic. No text, no logo, no watermark.

Goal: Create a vertical Japanese manga page about {argument name="character name" default="Faneria"}, a tiny silver-haired maid working in a luxurious mansion for a plush white rabbit master. The humor is that the maid's hairstyle is inconsistent across panels, sometimes twin tails and sometimes ponytail-like, while her face is covered by a plain beige rectangular privacy block in most close views.

Canvas: Tall smartphone comic page, approximately 2:3 aspect ratio, full color, high-detail anime illustration, thick black outer border, black manga gutters, slightly tilted panel borders, glossy digital painting with warm lighting.

Layout: Use exactly 5 panels arranged vertically: two small panels across the top row, one wide action panel in the upper middle, one wide dining panel below it, and one large dramatic night panel at the bottom.

Characters and repeated elements: Include exactly 2 main characters: 1 petite anime maid girl and 1 anthropomorphic white rabbit gentleman. The maid has pale silver-white hair, a black maid dress with white apron, white frilled headband, oversized black hair bows, black stockings, small shoes, and a cute chibi body proportion; place a flat beige rectangular censor block over her face in panels where her face is frontal or prominent. The rabbit is a soft white plush-like bunny with long ears, usually wearing dark formal clothes or pajamas, expressive closed smiling eyes, and refined aristocratic behavior.

Panel details: Panel 1, top left: in an ornate mansion room with mirrors and gold trim, the rabbit in a black butler-like suit stands with his back turned while the maid follows holding a clipboard; the maid's silver hair appears long and tied with black bows. Add a vertical white narration box reading 「旦那様のスケジュールは朝からビッシリです」. Panel 2, top right: bedroom scene with red curtains and antique furniture; the rabbit sits in a luxurious bed wearing blue-gray pajamas and a navy sleep mask pushed onto his forehead, sipping morning tea from a tray held by the maid; add sleepy sound effect 「ん〜…」 and a vertical narration box reading 「朝はまず美味しい紅茶でお目覚めになって頂きます」. Panel 3, middle wide action panel: grand hallway with chandeliers and red carpet; the maid, now clearly shown with twin-tail-like silver hair, uses a large vacuum cleaner or blower wand to blast away exactly 6 small black insect-like pests with a bright blue jet of sparkling light; add large blue sound effect 「シュバーン!!」 and a vertical narration box reading 「お屋敷を美しく保つのも大事な仕事です」. Panel 4, lower wide dining panel: candlelit banquet table packed with ornate dishes, including a large lobster platter, salads, roast meats, small plates, wine glasses, and a wine bottle; the maid presents a large dish while the rabbit happily eats with a fork and napkin; add pink chewing sound effect 「もぐもぐ」 and a vertical narration box reading 「ディナーは最高級、一日の労をねぎらうのです」. Panel 5, bottom large panel: nighttime exterior of the mansion under a full moon; the maid is in a dynamic battle pose holding a huge mechanical hammer or weapon, with silver hair flying dramatically and a beige face block; a large broken dark robot or monster lies in front with glowing red eyes, smoke puffs, sparks, and stars; in the mansion window the rabbit sleeps peacefully with 「Zzz」 above him; add red impact sound effect 「ドンッ!!」 and a vertical narration box reading 「24時間旦那様の安全を確保することが最優先」.

Visual style: Polished anime manga, cinematic lighting, ornate Victorian mansion interiors, rich browns, golds, burgundy curtains, sparkling highlights, dramatic action effects, cute chibi proportions mixed with detailed backgrounds, Japanese commercial webcomic look.

Text count and constraints: Include exactly 6 white vertical narration boxes with black Japanese text, exactly 4 sound effects including 「ん〜…」, 「シュバーン!!」, 「もぐもぐ」, and 「ドンッ!!」, plus one small 「Zzz」 sleep text. Do not add extra panels, extra main characters, logos, watermarks, or English captions.

Create a highly detailed fantasy anime illustration of an elegant young woman in a radiant crystal palace, wearing an emerald green fantasy gown that looks like transparent gemstones and flowing aurora glass. She has {argument name="hair color" default="pale silvery mint-blonde"} long wind-swept hair, styled with an elaborate jeweled floral hair ornament made of emeralds, diamonds, and iridescent crystal petals. Her face is softly lit and partially turned in a graceful three-quarter pose, with a dreamy, ethereal atmosphere. The dress is an off-shoulder, fitted bodice gown in {argument name="dress color" default="emerald green"}, made from translucent layered organza, crystal-like fabric, sparkling beadwork, gold filigree, and large faceted emerald gems; include one prominent jeweled flower at the waist and delicate gem chains across the neckline and bodice. The skirt and sleeves should billow dramatically around her like liquid glass, with rainbow prism reflections, star-like glitter, and refracted light. Place her in a luminous cathedral-like crystal garden with tall arched columns, ornate glass architecture, hanging prisms, crystal chandeliers, white lilies, and sparkling foliage. Use strong warm sunlight streaming from the upper left, creating lens flare, glowing backlight, bokeh, and rainbow caustics across the scene. Style: ultra-detailed Japanese fantasy art, romantic shoujo elegance, luminous painterly rendering, delicate linework, high sparkle density, soft pastel highlights, transparent gem textures, magical luxury, vertical composition, 4:5 aspect ratio. Keep the image focused on one full upper-body figure, no text, no watermark, no extra characters.

Goal: Create a dark-mode mobile chat interface screenshot showing an AI image generation result for an article illustration.

Canvas: Portrait smartphone screenshot, 896×768 style crop, black background, with a large rounded dark-gray prompt bubble near the top and a generated image preview card below it. Leave generous black margins around both cards.

Top prompt bubble: One charcoal-gray rounded rectangle aligned toward the upper right, partly cropped at the very top. It contains white and gray prompt text in English, including the visible lines: “slight angle, looking at the café window with a slightly uncertain posture”, a standalone dash, “hands in pockets or arms crossed. Warm light glowing”, and a dimmer line “from inside the café.” At the bottom of this bubble, include the Japanese UI text “表示を増やす” with a small downward chevron.

Generated image card: One large rounded-rectangle image preview beneath the prompt bubble, with strongly rounded corners. Inside it, paint a soft, warm, nostalgic watercolor-style illustration of {argument name="scene subject" default="a young woman standing alone outside a small neighborhood café at dusk"}. The woman is viewed from behind at a slight angle, looking into the café window with a hesitant, uncertain posture. She has dark hair tied in a low bun, wears a long beige coat over a dark ankle-length skirt, white shoes, and carries a cream tote bag over one shoulder; her hands are in her coat pockets. The café façade has large glass windows, weathered wooden double doors, warm amber interior lighting, pendant lamps, shelves, a coffee counter, faint customers or staff silhouettes, and a chalkboard menu inside. The street is quiet and slightly rainy or overcast, with utility poles, overhead wires, narrow sidewalk, muted buildings receding into the distance, and a pale evening sky. Add exactly 8 visible plant containers around the storefront: 2 taller shrubs flanking the window/door area, 3 medium pots near the right side, 1 small white pot near the bench, 1 planter near the sidewalk sign, and 1 low pot by the step. Add exactly 2 street-side objects in front of the café: a small wooden bench and a black A-frame chalkboard sign.

Overlay UI on image card: Add exactly 2 circular translucent buttons on top of the image: a bottom-left button labeled “編集” and a bottom-right share/upload icon button. Keep the buttons subtle, light gray, and semi-transparent.

Visual style: Gentle Japanese urban slice-of-life illustration, muted beige and gray palette, warm café glow, painterly watercolor texture, soft edges, slightly grainy paper feel, atmospheric and cute but realistic. The image preview should look like a finished generated illustration embedded in a chat app, not a full-bleed artwork.

Constraints: Use only the visible UI elements described: 1 top prompt bubble, 1 generated image card, 2 overlay buttons, and the Japanese text labels exactly as written. No extra captions, no watermark, no app logo.

Create a painterly nighttime city street scene in a cinematic square composition. The focal point is a small modern convenience store on the right side of the frame, glowing warmly through large glass windows and sliding doors, with shelves of colorful products visible inside. Above the entrance is a blue fascia sign with a thin hot-pink stripe and the readable text “24 HOURS” in white. Place exactly 2 vending machines just outside the store near the entrance, plus exactly 2 dark trash bins beside them. The street is empty of people and cars, quiet after rain, with wet asphalt and puddles reflecting the store’s orange-pink light and the deep blue sky. Include utility poles, overhead power lines, dim streetlights, low buildings fading into mist on the left, and one large dark pole in the foreground on the right edge. Use an impressionistic digital oil-painting style with visible brush texture, rich ultramarine and indigo shadows, neon magenta accents, warm peach interior lighting, soft haze, and dramatic reflections. Keep the mood melancholic, calm, and atmospheric; no extra text besides the store sign.

Create a polished anime-style double-exposure illustration of a young woman in a white Japanese kimono under a transparent umbrella, viewed from a low rear three-quarter angle as she turns her head back over her shoulder. The composition should be a vertical portrait with bright summer daylight, a vivid blue sky, and delicate cherry blossom branches visually blended through the umbrella canopy as a double-exposure effect. The subject wears a glossy white kimono with subtle embroidered floral patterns, a red inner collar, and a large red obi bow at the back with cream-and-gold decorative fabric and red flower motifs. She has {argument name="hair color" default="dark blue-black with ash-brown curled ends"} hair styled in an elegant updo with loose curls, decorated with a large white folded hair ornament, one vivid red rose, and one small gold circular accent. Add a dangling pearl earring on the visible ear. Her visible hand delicately holds the black umbrella handle behind her, with neat red-painted nails. Use luminous painterly rendering, crisp anime linework, soft gradients, reflective highlights on the kimono fabric, and high detail in the hair, folds, blossoms, and umbrella ribs. Keep the face mostly turned away and serene, emphasizing the graceful pose, the red-white-blue color palette, and the dreamy double-exposure umbrella filled with {argument name="seasonal flowers" default="pale pink cherry blossoms"}. Avoid text, logos, watermarks, extra characters, or modern city details.

Create a dramatic square fantasy portrait titled around the theme of {argument name="theme" default="Light and Shadow"}: a female knight shown from the torso up, centered, with her body split into two opposing halves. The left half is luminous and fiery, with wild flowing {argument name="light side hair color" default="golden blonde"} hair blown outward, warm cream background, orange and magenta paint splashes, and bright high-contrast highlights. The right half is dark and gothic, with deep {argument name="shadow side hair color" default="black-blue"} hair, navy ink washes, a gloomy castle silhouette in the background, and a vertical sword standing beside her shoulder. She wears ornate engraved black-and-silver medieval armor with a decorative cross-like chest ornament, rendered with gritty ink texture, halftone dots, scratches, splatters, and mixed-media brushwork. Add exactly 2 opaque rectangular face-obscuring blocks: one large warm beige-to-dark-brown square centered over the left/middle face, and one dark navy-to-gray vertical rectangle over the right face, slightly smaller and positioned higher-right. Use a dynamic fashion-illustration/comic-poster style with explosive paint splashes, rough paper texture, sharp contrast, and no readable text or watermark.

Create a polished anime-style illustration of a young Japanese-style maid resting in a traditional tatami room on a bright summer day. The scene features exactly one seated maid, exactly one round wooden tea tray, exactly one clear glass cup filled with green tea, exactly one hanging glass wind chime with a blue paper strip, exactly one wall scroll painting, exactly one ceramic vase with greenery, and exactly one white handwritten artist signature reading “Okamata” in the bottom-right corner. The maid wears a hybrid kimono maid outfit: an olive-green kimono top with a red-and-white collar, a white frilled maid apron with ruffled shoulder straps and waist sash, a black-and-gray checkered pleated skirt, white thigh-high stockings, and a white maid headband. She is seated on tatami with one hand braced on the floor and the other resting near her skirt, posed as if taking a quiet break in the breeze. Her hair is {argument name="hair color" default="short silver-white"}, softly lit and slightly tousled. The face is intentionally covered by a plain opaque beige rectangular censor block with no facial features. The setting is a traditional Japanese room with wooden beams, sliding shoji doors opened to a sunlit veranda, glossy wooden floorboards, garden railing, green trees, distant mountains, and a small temple-like roof visible outside under a vivid blue sky with scattered clouds. Use warm sunlight, strong natural highlights, soft shadows, detailed fabric folds, crisp linework, glossy anime rendering, and a calm summer atmosphere. Customize the maid outfit color as {argument name="kimono color" default="olive green"}, the drink as {argument name="tea type" default="green tea"}, the outdoor season as {argument name="season" default="summer"}, and the signature text as {argument name="signature text" default="Okamata"}. Keep the composition vertical, waist-up to knees visible, with the character centered and the open veranda framing the bright landscape behind her.

Goal: Create a vertical smartphone screenshot of a romantic comedy anime manga page about a sudden outfit change during a date, with the characters' faces intentionally covered by soft rectangular privacy blurs.

Canvas: Tall 9:16 mobile screenshot, approximately 625 x 1080, with a black app viewer header at the top. Header includes a back arrow on the left, centered white Japanese title text {argument name="header title" default="服装チェンジ(今だけ)(1/1)"}, and two small white interface icons on the right. Add two semi-transparent circular navigation buttons with left and right chevrons near the bottom center.

Layout: Use exactly 6 tilted comic panels separated by thick black gutters. Panel 1, top left: close-up of a petite blonde young woman in a purple-and-white maid-like dress, one finger near her lips, embarrassed. Panel 2, top right: handsome dark-haired young man in a black butler tuxedo standing on a European street. Panel 3, middle left: the same butler man scratching the back of his head awkwardly. Panel 4, middle right: close-up of the blonde woman blushing heavily, overwhelmed. Panel 5, bottom large panel spanning most of the width: the blonde woman has fallen forward onto the cobblestone street, flustered, skirt and translucent sleeves spread dramatically. Panel 6, bottom right narrow panel: partial view of the butler man reaching out in surprise.

Characters: The woman is {argument name="female character" default="a delicate blonde anime girl with long wavy hair, side braids, purple flower hair ornaments, violet tassel earrings, flushed cheeks, and a frilly white-and-purple maid dress with translucent lace sleeves"}. The man is {argument name="male character" default="a tall handsome anime man with tousled black hair, a black butler suit, gold trim, white ruffled shirt, black bow tie, lapel chain, and a decorative brooch"}. Cover the visible faces with exactly 5 soft rectangular censor blurs: one over the woman's face in panel 1, one over the man's face in panel 2, one over the man's face in panel 3, one over the woman's face in panel 4, and one over the partial man's face in panel 6; leave the fallen woman's closed-eyed face in panel 5 visible.

Text content: Use exactly 10 speech bubbles and preserve the Japanese dialogue in this set: {argument name="Japanese dialogue text" default="どうして… / いつもはメイド服なのに…！ / おーいミワどうしたんだ？ / てか貸してくれるか？ / ？ / そんなに驚くことか？ / あーたまにはいいだろ？ / かっこよすぎる…っ / ずるいよこんなの反則でしょー!!! / おい、ミワ!?"}. Add exactly 4 sound-effect or reaction texts: {argument name="sound effects" default="はなそそや… / きゅううん…♡ / はぁっ / ガクッ!!"}.

Visual style: Polished high-detail AI anime illustration, soft romantic lighting, pastel highlights, sparkles, shallow depth of field, ornate European city street background, glossy manga rendering, dramatic diagonal panel composition, expressive rom-com pacing. Keep the image vertical, dense, and screenshot-like; do not add extra panels, extra characters, watermarks, or additional text.

Create a full-body anime character illustration of {argument name="character name" default="a mischievous demon schoolgirl"} standing confidently in the foreground outside a modern school building on a sunny afternoon. She has {argument name="hair color" default="short black bob-cut hair with messy bangs"}, tan skin, sharp half-lidded eyes, a sly smirk, rosy cheeks, and exactly 2 small glossy black demon horns on top of her head. Her outfit is a dark navy Japanese-inspired school uniform: white short-sleeve collared shirt, navy sleeveless sweater vest, large red bow at the chest, black pleated mini skirt, sheer black tights, and chunky black platform loafers. Add exactly 4 visible accessories on her body: a black choker with a heart-shaped ring pendant, a red scrunchie bracelet on one wrist, black nail polish, and a large black shoulder bag. The shoulder bag should hang from her right shoulder and include exactly 5 decorations: a cute black cat face patch, a small red heart, 2 star charms, and a dangling chain strap detail. Pose her with one hand on her hip and the other gripping the bag strap, legs crossed slightly as if walking toward the viewer.

Set the scene in front of a stylized school campus with a warm cartoon/anime look inspired by early-2000s Western animation blended with polished Japanese anime. The background contains exactly 4 secondary students: 2 girls sitting on a bench at left, one orange-haired girl drinking from a cup and one dark-haired girl reading a purple book; and 2 boys standing near the entrance at right, one orange-haired boy in a red jersey holding a skateboard and one black-haired boy in a dark shirt with green pants. Include a large school clock above the entrance and a black-and-white sign reading {argument name="school sign text" default="CN"} with smaller text below reading "CARTOON NETWORK". Add leafy green trees on the left, blue sky with a few soft white clouds, glass school windows, concrete steps, sidewalk tiles, scattered leaves, and graffiti on a low wall with exactly 3 marks: a crown, a heart, and a star.

Use vibrant cel-shaded lighting, crisp line art, saturated colors, exaggerated expressive proportions, and dramatic late-afternoon shadows. The camera angle is slightly low and centered, emphasizing the main character as tall and charismatic. Vertical portrait composition, full body visible from head to shoes, highly detailed but clean, no extra characters, no watermark, no photorealism.

Create an ultra-realistic vertical beach portrait of a graceful South Asian woman sitting alone on pale sand beside turquoise ocean waves under bright white midday sunlight. She wears an elegant {argument name="outfit color" default="cream"} traditional sari with premium silk texture, subtle gold embroidery, ornate border detailing, and a matching short-sleeved blouse. She has long flowing {argument name="hair color" default="dark brown"} wavy hair and a delicate floral crown made of small white baby's-breath flowers with soft blush pink accents. Pose her seated in a relaxed side-facing position with the sari draped across her lap, one arm resting across her knees and one hand gently placed near the sand, conveying serene romantic elegance. The background shows a calm tropical shoreline with foamy waves, a blue-green sea, a clear pale blue sky, and distant rocky cliffs softly blurred by shallow depth of field. Use natural skin tones, luxury editorial fashion photography, soft highlights, crisp fabric detail, gentle wind in the hair and sari, realistic shadows on the sand, and a high-end dreamy coastal wedding aesthetic. Compose as a 2:3 vertical frame with the woman centered slightly left, full body visible from head to flowing sari hem, no text, no watermark, no extra people.

Create an ultra-realistic vertical cinematic fashion editorial portrait of an {argument name="age and subject" default="18-year-old young woman"} standing indoors in a warm minimalist apartment or studio. She is framed from upper thighs to head, centered, with a relaxed confident pose: one arm down with a bent elbow, the other hand resting lightly near her waist. She wears an oversized rust-orange button-up shirt with a soft cotton/linen texture, open collar, visible front buttons, dropped shoulders, and sleeves rolled up to the forearms. Add exactly 3 visible accessories: clear-framed eyeglasses resting on top of her head, a small stud earring on one ear, and a silver wristwatch on her right wrist. Her hair is a modern layered {argument name="hair style" default="medium-length wolf cut"} with soft volume, natural texture, tousled outward layers, and {argument name="hair color" default="dark brown with subtle warm highlights"}. Keep the face area intentionally covered by a centered vertical rectangular privacy block in a medium warm brown color, obscuring all facial features from forehead to chin while leaving the hair, ears, neck, and outfit visible. Use warm natural window light, shallow depth of field, creamy beige background blur, high-end editorial photography, realistic skin on visible neck and hands, subtle fabric wrinkles, soft shadows, 85mm lens look, f/1.8 bokeh, natural color grading, no text, no watermark, no extra people.

Goal: Create a surreal Japanese advertising collage for {argument name="brand" default="NISSIN"} cup noodles, mixing a snowy pastoral landscape, internet culture, AI imagery, and absurd product placement.

Canvas: Wide horizontal banner, approximately 16:9, bright cyan winter sky and snowfield background, glossy commercial-composite style with photorealistic objects, soft glow effects, and exaggerated ad-like color saturation.

Main subject: Center-left foreground, a young blonde woman in a white lace top faces the viewer. Her face is intentionally obscured by a large flat tan square censor block. She holds an open silver laptop/tablet in her left hand and presents a cup noodle product on her right palm. The cup is labeled as a milk seafood noodle flavor, with colorful Japanese packaging and a creamy instant noodle lid. A small green mascot-like robot/alien with two antennae stands on or behind the noodle cup; its face/body is partly blurred by a small green square.

Text and branding: Top-left corner shows a red {argument name="logo text" default="NISSIN"} badge with small Japanese text beneath it reading 日清商品. Place a large white Twitter bird icon nearby in the upper-left sky. Bottom-left to lower-center has oversized rounded Japanese headline text {argument name="headline text" default="あぁ、チーズ感。"} in glossy orange-yellow 3D lettering with a soft shadow, spanning across the snow and foreground.

Background elements: Create a snowy landscape with white trees on the left, rolling snow hills, fences, and a farmhouse or barn on the right. In the upper-right sky, show a huge glowing translucent blue digital brain made of neon circuitry. Add sparkling data particles and faint interface marks around it.

Counted floating objects: Include exactly 5 floating balloon/orb elements in the sky: 1 hot-air balloon labeled “internet,” 1 smaller hot-air balloon labeled “Orotier,” 1 large hot-air balloon labeled “SNS” twice, 1 glowing glass orb labeled “Digital,” and 1 small orange-yellow hot-air balloon labeled “Oratier.”

Counted foreground and midground objects: Include exactly 3 food/drink items on the left snow beside the woman: 1 wine glass with red wine, 1 green wine bottle, and 1 cup noodle held in the woman’s hand. Include exactly 2 pigs lying on the snow near the lower-right center. Include exactly 1 panda standing at far right holding bamboo, with extra bamboo leaves on the snow. Include exactly 2 children in the snowy midground using devices: one child in a red striped shirt holding a tablet while running/kneeling near the pigs, and one child near the right farmhouse holding a laptop/tablet. Both children’s faces are blurred or covered. Include exactly 1 dark glossy futuristic vehicle or sled partially buried in snow near the children, with blue glowing highlights.

Style constraints: Make the image intentionally chaotic and meme-like, as if testing text rendering in an AI-generated Japanese ad. Preserve readable text where specified, but allow minor imperfect AI typography on small labels. Use bright daylight, crisp snow, commercial product lighting, and seamless but uncanny photomontage compositing. Do not add extra people, animals, logos, or headline text beyond the specified counts.

Create a high-fashion street style editorial portrait of a stylish young woman standing outdoors in front of a softly blurred modern city building. She has {argument name="hair color" default="dark brown"} hair styled into one long, thick French side braid draped over her right shoulder, with a few loose face-framing strands. Her face is intentionally obscured by a centered square mosaic blur, but black designer sunglasses are still partially visible at the sides. She wears a premium oversized {argument name="t-shirt color" default="beige"} short-sleeve T-shirt tucked into high-waisted tailored {argument name="trouser color" default="beige"} trousers with belt loops and front pleats. Include exactly two visible jewelry/accessory pieces: a minimalist thick gold chain necklace around her neck and a luxury silver wristwatch on the wrist of the hand raised to touch the braid. Pose her from mid-thigh upward, body angled slightly left, one arm relaxed down and the other bent with the hand near the braid. Use natural daylight, shallow depth of field, realistic fabric texture, muted neutral tones, luxury minimalist styling, editorial fashion photography, 1:1 square composition, crisp subject focus, soft urban bokeh background, no text, no watermark.