GPT Image 2 prompts

Create a delicate vertical romantic anime-style watercolor illustration of {argument name="character name" default="a young bride"} and {argument name="partner" default="a pale-haired young man"} standing very close in a sunlit interior, as if about to kiss or sharing a private breath. The bride is shown from behind and slightly in profile, with dark brown hair gathered into a loose low bun, many wispy strands, and an ornate white floral hair ornament with dangling pearl-like beads; she wears a sheer ivory wedding kimono or bridal robe with soft translucent fabric and fine sketched folds. The partner appears on the upper right edge, only partially visible with light blond or silver hair and a gentle lowered face. Cover the central face-contact area with one large opaque square censor block in warm beige, positioned over both faces and spanning the center-right of the image. Use a muted cream, beige, ivory, and pale gold palette, very soft backlighting, dusty sunbeams, faint floral wallpaper or window reflections in the background, and a quiet nostalgic atmosphere suggesting restrained longing and tenderness. The style should be fine-lined, airy, painterly, highly detailed in hair strands and accessories, with low contrast, translucent washes, and no visible text or watermark.

Create a realistic vertical smartphone photo in the style of a bright Japanese summer beverage commercial, showing {argument name="number of students" default="three"} cheerful high school girls cleaning an outdoor school swimming pool deck on a sunny day. The scene is candid and ultra-realistic, with natural daylight, blue sky, soft white clouds, wet reflective concrete, sparkling water spray, and a fresh youthful atmosphere. All students wear matching Japanese school uniforms: white short-sleeve button-up shirts, loosened navy striped neckties, blue plaid pleated skirts, and they are barefoot on the wet poolside. Count exactly three girls: the center girl is closest to camera, leaning forward while pushing a long-handled deck brush toward the viewer, with a short dark bob haircut and rolled-up sleeves; the left girl stands near a chain-link fence holding a running blue hose, spraying water onto the ground, with a navy school bag hanging from her shoulder and small cute keychains attached; the right girl stands upright with one hand on her waist, holding a cup of iced bubble tea with a straw in the other hand, long dark hair tied back, relaxed confident pose. Include exactly three main cleaning/drink props: one deck brush, one blue hose spraying water, and one bubble tea cup. Add one bright blue plastic bucket at the lower right edge. Background details: turquoise chain-link fence along the left side, school buildings and pool structures in the distance, bright blue swimming pool visible behind the girls. Composition: low eye-level to slightly low-angle full-body shot, 9:16 vertical framing, center subject dominant, left and right girls balancing the composition, shallow but realistic depth of field. Visual style: photorealistic Japanese commercial photography, crisp skin and fabric texture, glossy wet floor reflections, high dynamic range, clean color grading, summery blue-and-white palette, no text, no watermark, no extra people. Faces should be naturally attractive and realistic, with the option to customize {argument name="center hairstyle" default="short dark bob"}, {argument name="left hairstyle" default="dark ponytail"}, {argument name="right hairstyle" default="long dark hair tied back"}, and {argument name="setting" default="outdoor Japanese high school poolside"}.

Goal: Create a cute vertical four-panel manga comic about a sheep girl and her tiny sheep encountering a magical portal and a cat.

Canvas: Portrait comic page, 768 x 1152 style, warm pastel colors, thin dark brown panel borders, soft cel-shaded anime illustration, cozy slice-of-life fantasy mood. Arrange exactly 4 horizontal panels stacked vertically with narrow white gutters.

Main characters: A chibi sheep-horned girl named {argument name="character name" default="HITSURO"}, with cream-white bobbed hair, curled ram horns on both sides, small pointed elf ears, large brown eyes, a white blouse, pale pink apron with pocket and bow, dark pleated skirt, knee socks, and soft shoes. Her face in the first panel is intentionally obscured by a plain beige rectangular censor block. A tiny fluffy white sheep companion with round black eyes, pink cheeks, small floppy ears, and a stubby tail appears in every panel. A calico cat appears in panels 3 and 4, white with orange and dark brown patches, round eyes, and a playful expression.

Layout and story: Use exactly 4 panels.
1. Panel 1: Cozy room interior with wooden floor, bookshelf, bottles, potted plants, framed wall art, curtain, and window. The sheep girl stands on the right facing away/sideways, her face covered by a beige rectangle. The tiny sheep stands on the floor to her left, trembling slightly with motion marks. Add a tall vertical Japanese speech bubble on the right containing {argument name="panel 1 dialogue" default="…揺れてる…ね"}.
2. Panel 2: The wooden floor has cracked open into a glowing jagged hole filled with a purple starry galaxy. The tiny sheep is partly emerging from or leaning over the hole, sparkling. The sheep girl crouches on the right, concerned, reaching toward it. Add a vertical speech bubble on the right containing {argument name="panel 2 dialogue" default="ひつ郎、近づいちゃだめ"}.
3. Panel 3: A vertical oval purple portal opens on the left, filled with a starry night sky and a glowing yellow star. A calico cat leaps out of the portal toward the room, with speed lines radiating dramatically. The tiny sheep stands near the center-right, startled and shaking. The sheep girl stands on the far right, surprised. Include exactly 2 speech bubbles: a small bubble near the cat saying {argument name="cat sound" default="ニャッ"}, and a tall vertical bubble on the right saying {argument name="panel 3 dialogue" default="…どこの子？"}.
4. Panel 4: The portal remains glowing on the left side of the cozy room. The tiny sheep floats or jumps out from the portal while the sheep girl kneels on the right and hugs the calico cat protectively. The cat looks pleased with closed eyes. Small glowing stars sparkle around them, and a few scattered small objects lie on the floor, such as a spool, coin-like circles, a small cloth, and a tiny container. Add a tall vertical speech bubble on the right containing Japanese text meaning “Hey, stop that,” using 「こら、やめなさい」.

Visual style: Soft kawaii manga, rounded character proportions, expressive eyes, cream and beige room palette, gentle brown line art, pastel purple magical glow, subtle sparkles, clean readable speech bubbles with vertical Japanese text. Keep the page wholesome and humorous.

Constraints: Include exactly 4 panels, exactly 3 recurring characters total across the story: the sheep girl, the tiny sheep, and the calico cat. Count the major repeated elements as follows: 4 speech bubbles total on the right side, 1 small cat sound bubble, 2 magical portal/hole events before the final panel, and 1 beige face-obscuring rectangle in the first panel only. Do not add extra panels, extra characters, watermarks, logos, or narration boxes.

Create an ultra-wide cinematic anime fantasy battle scene in a dark ruined volcanic battlefield at night, aspect ratio about 2.4:1. Show exactly two female warriors clashing in midair among jagged black rocks, smoke, sparks, embers, and distant fires. On the left, {argument name="left warrior" default="a fierce red-haired fire knight"} with long windblown crimson hair, ornate gold plate armor, a flowing burgundy cape and skirt cloth, lunging forward with a glowing orange fire sword and a swirling flame aura behind her. On the right, {argument name="right warrior" default="a silver-haired ice warrior"} with very long flowing silver-white hair, glossy black-and-white battle armor with blue crystal accents, crouching and bracing with one arm extended; place a flat dark gray square censor block over her face. Between them, render a large circular blue magic shield sigil made of intricate glowing runes and geometric rings, struck by the fire blade with bright orange sparks at the impact point. Add exactly two elemental effects: orange fire around the left warrior and blue ice crystals around the right warrior, including sharp translucent shards bursting outward behind her. Use high-detail Japanese anime key visual style, dramatic rim lighting, dynamic perspective, sharp linework, painterly rendering, intense motion, high contrast orange-and-cyan color grading, cinematic depth of field, no text, no watermark.

Create a full-body vertical fashion portrait of {argument name="character name" default="an elegant gothic lolita young woman"} standing in a dark Victorian room, facing the camera with a graceful doll-like pose, one hand lightly holding out the skirt and the other relaxed at her side. She wears an elaborate black-and-royal-blue aristocratic gothic lolita dress with gold embroidery, a fitted corset bodice with crisscross gold lacing, long flared sleeves, white lace cuffs, a high white ruffled collar, a large satin blue bow at the neck with a jeweled brooch, and a second large blue waist bow with dangling gold tassels. The skirt is dramatic and layered: black outer overskirt with ornate gold filigree trim, deep royal-blue ruffle tiers, translucent pale-blue lace layers, and white lace edging, forming a high-low bell silhouette. Add black floral lace thigh-high stockings, black glossy platform Mary Jane shoes with buckles, and delicate gold jewelry. Her hair is {argument name="hair color" default="dark brown wavy twin-tail hair"}, decorated with exactly 4 visible head accessories: a small tilted black top hat with a blue band, a black lace veil, a cluster of blue roses, and gold filigree hair ornaments. Use a refined color palette of {argument name="main colors" default="black, deep royal blue, antique gold, and white lace"}. The setting is an ornate gothic Victorian interior with exactly 6 prominent decor elements: a carved black-and-gold cabinet on the left, lit silver candelabras with candles, dark blue roses in a vase, heavy navy velvet curtains, a tall arched leaded-glass window on the right, and a small black-and-gold side table holding a blue rose. Add a patterned blue-and-gold rug on a dark floor. Lighting should be moody candlelit chiaroscuro with cool window light, rich shadows, luxurious textures, ultra-detailed fabric, realistic photography, sharp full-body composition, vertical 9:16 framing. Include the visible white watermark text {argument name="watermark text" default="Pollo.ai"} in the top-right corner. Do not add extra people, extra text, modern objects, or a visible face blur; keep the face naturally rendered if possible.

Create a polished anime fantasy illustration of {argument name="character name" default="a graceful pink-haired noble girl"} standing on a moonlit castle balcony at night, shown from mid-thigh up in a vertical composition. She has short fluffy bobbed {argument name="hair color" default="pink"} hair with long side bangs partly covering one eye, large ruby-pink eyes, a soft shy smile, pale skin, and delicate features. Add a white flower crown across the top of her hair, a purple flower-shaped hair ornament on the right side, and a small cross hairpin on the left. Dress her in an ornate gothic lolita outfit in {argument name="dress colors" default="black, deep magenta, and rose pink"}: fitted bodice, puffed skirt with layered ruffles, white frilled blouse details, bow at the collar, capelet over the shoulders, rose brooch at the waist, fine gold chain decorations, embroidered floral trim, white cuffs, and frilly underskirt. Her left hand rests gently near her chest, while her right arm reaches outward toward the viewer with an open palm, elegant but slightly stylized anime fingers. The setting is a romantic European castle terrace with stone balustrades, blooming rose bushes, tall arched windows glowing warm yellow, distant gothic towers, and a large full moon in a deep navy starry sky. Surround her with swirling magical ribbons of pink light and many floating rose petals, including foreground petals blurred for depth. Use cinematic moonlight mixed with warm window glow, soft rim lighting, high detail, glossy anime rendering, dreamy atmosphere, rich magenta highlights, shallow depth of field, and a clean high-resolution illustration style. No text, no watermark, no modern objects.

Goal: Create a clean vector icon set for business presentation illustrations, suitable for slide decks and documents.

Canvas: Wide 16:9 white background, evenly spaced grid layout with 40 discrete icons arranged in 5 rows and 8 columns. Use generous margins and consistent icon sizing.

Visual style: Flat modern corporate vector style with rounded shapes, minimal shading, crisp edges, and a friendly SaaS/business look. Use a limited palette of {argument name="primary blue" default="bright blue"}, {argument name="accent green" default="fresh green"}, {argument name="accent orange" default="vivid orange"}, dark navy outlines, light gray UI details, and occasional yellow. No text labels, no watermark, no realistic texture.

Icon contents: Include exactly 40 icons in this order, left to right by row:
Row 1: 1) three people meeting at a table with speech bubbles, 2) bar chart with upward growth arrow, 3) line graph with three colored points and rising arrow, 4) three vertical upward arrows in blue green orange, 5) presentation screen with pie chart and pointer, 6) laptop with dashboard cards, 7) tablet with colored dashboard blocks.
Row 2: 8) kanban board with three columns and colored task cards, 9) glowing light bulb idea icon, 10) target with arrow hitting bullseye, 11) handshake with blue and green sleeves, 12) monthly calendar grid with one orange date, 13) document page with blue folded corner and lines, 14) checklist document with green check marks.
Row 3: 15) cloud upload icon with upward arrow, 16) dark blue gear settings icon, 17) magnifying glass search icon, 18) blue envelope email icon, 19) video conference monitor with four participants, 20) gold trophy award icon, 21) four-piece puzzle in blue green orange.
Row 4: 22) green circle check mark, 23) pie chart in blue green orange, 24) organization chart with connected colored nodes, 25) thumbs-up approval icon, 26) team/group icon with three people, 27) bar chart with rising green arrow, 28) green shield with white check mark.
Row 5: 29) money bag with dollar sign and stacked coins, 30) calculator with orange key, 31) analog clock, 32) yellow file folder, 33) overlapping chat bubbles with ellipsis, 34) dark blue briefcase with yellow clasp, 35) blue globe wireframe.

Layout constraints: Maintain a balanced grid and consistent visual weight; leave open white space between icons; keep all icons centered in their cells; avoid adding captions or extra elements. The final image should feel like a reusable {argument name="asset type" default="business icon library"} for {argument name="use case" default="presentation materials"}.

Create a dramatic square anime illustration of exactly two teenage girls in dynamic motion, seen from a low, tilted angle as if they are falling, floating, or dancing through an abstract storm of ink. The first girl has {argument name="light-haired girl hair color" default="pale blonde hair"} tied in a messy high ponytail with loose strands covering part of her face; she wears a white Japanese sailor school uniform with dark collar trim, a tied neckerchief, a pleated skirt, white socks, and black shoes. She is positioned on the upper-right side, body angled diagonally, one knee bent, one arm extended back, looking downward with a soft serious expression. The second girl has {argument name="dark-haired girl hair color" default="long black hair"} flowing wildly across the left side; her face is mostly hidden in shadow and hair. She wears a dark, tattered school-uniform-like dress or blouse and skirt, rendered almost as a black silhouette, with deep crimson tones around the lower fabric. Place her on the lower-left side, reaching or drifting toward the blonde girl, creating an intimate but tense two-character composition. Use a gritty monochrome palette of warm ivory, charcoal black, sepia brown, and muted blood-red accents. Style it as expressive Japanese anime concept art mixed with ink wash, scratchy pen lines, watercolor stains, splattered black ink, rough brush textures, distressed paper grain, and high-contrast chiaroscuro. Background is mostly off-white empty space with explosive black and brown ink splashes concentrated in the upper-right and lower-left corners, diagonal motion streaks, and scattered speckles. Emphasize flowing hair, pleated fabric movement, dramatic shadows, delicate faces, painterly anatomy, and a cinematic, melancholic atmosphere. No text, no border, no watermark.

Goal: Create a vertical Japanese awareness poster for {argument name="observance date" default="May 14"}, Thermometer Day, combining a clean educational infographic with a realistic model wearing a themed T-shirt.

Canvas: Tall vertical social-media poster, 9:16 aspect ratio, pale ice-blue background, soft clinical lighting, clean magazine-style layout, blue and red accent palette.

Main subject: On the right half, place a realistic young Japanese woman with a neat chin-length black bob haircut, facing forward from upper thighs up. Her face is intentionally obscured by a plain rectangular blur/mosaic block. She wears a fitted short-sleeve gradient T-shirt fading from sky blue on the left to pink/red on the right. The shirt text reads exactly: “5/14” on the first line, “温度計の日” large below, and “ドイツの物理学者 ファーレンハイトの誕生日” in small text underneath. She holds a white digital thermometer delicately near her shoulder in her right hand.

Top section: On the left, large Japanese headline text reads “5月14日は” and then “温度計の日”, with “温度計” in dark blue and “の日” in red. Add a small line-art thermometer icon above/right of the headline with “°C”. Subtitle below: “— 暮らしを支える、温度のものさし —”. Add a short explanatory paragraph in small Japanese text describing that May 14 is Thermometer Day, related to the German physicist Fahrenheit and the importance of temperature measurement for science, personal life, safety, and comfort.

Decorative icons: Include exactly 3 pale blue scientific/weather icons in the upper-middle left background: 1 vertical thermometer icon, 1 globe icon, and 1 laboratory flask icon.

Middle infographic section: Add a heading “温度計が活躍する場面” with a yellow underline. Beneath it, create exactly 3 rounded rectangular cards in a row, each with a blue outline, a small icon, a bold Japanese title, explanatory text, and a photo thumbnail:
1. Heart icon, title “健康管理に”, text about measuring body temperature and noticing changes early, thumbnail of a digital thermometer on a soft towel showing 36.5.
2. Cooking pot icon, title “料理・暮らしに”, text about cooking temperature control and bath temperature adjustment, thumbnail of a thermometer inserted into a steaming cooking pot.
3. Lab flask icon, title “研究・産業に”, text about accurate temperature measurement supporting technology and products, thumbnail of glass beakers/test tubes with a pipette.

Bottom section: Add a closing heading “温度を知ることは、未来を守ること” in dark blue, a thin horizontal rule, two lines of smaller Japanese explanatory text about climate awareness and protecting life and Earth’s future, and a final red call-to-action line: “今日は温度計に、ちょっと注目してみませんか?” Place a faint green-blue Earth-with-sprout icon at the bottom right.

Visual style: Clean Japanese public-information poster, realistic photography for the woman and card thumbnails, flat minimalist icons, soft shadows, rounded card corners, plenty of white space, professional print layout, no watermark, no extra logos. Use crisp readable Japanese typography and keep the exact visible section count: one top headline section, one model figure, three decorative icons, three application cards, and one bottom message section.

Goal: Create an anime character reference sheet featuring {argument name="character name" default="a young shrine maiden girl"} in a consistent red-and-white miko outfit, shown as a collection of full-body poses and expression variations on a clean white background.

Canvas: Square 1:1 image, high-resolution, white background with a very faint pale-gray grid and thin rectangular panel divisions like a character design board. Use crisp clean anime line art, soft cel shading, polished digital illustration, and consistent proportions across all versions.

Layout: Arrange exactly 15 full-body character figures in a loose 5-column by 3-row reference-sheet composition. The central figure is the largest and tallest, placed in the middle, front-facing, occupying most of the vertical height. The other 14 figures are smaller surrounding pose variations, each inside its own faint panel. Keep generous white space and avoid any text labels.

Subject details: The character is a petite teenage anime girl with very long straight {argument name="hair color" default="black"} hair reaching past the waist, blunt bangs, small side locks, pale skin, and red eyes when visible. She wears a traditional shrine maiden outfit: white kimono-style top with wide sleeves, red trim and red stitch-like sleeve markings, red hakama skirt with deep pleats, a large red waist bow/sash, white tabi socks, and red-and-white sandals. The outfit should remain consistent in every pose.

Pose and expression count: Include exactly 15 discrete figures: 1 large central front-facing neutral full-body pose with a softly blurred/featureless face; 2 top-left jumping pose with clenched fists and blurred face; 3 top-second standing pose with one hand on hip and blurred face; 4 top-center close-up standing pose with blurred face; 5 top-fourth small front-facing neutral pose with visible calm face; 6 top-right shy standing pose with clenched fists and blurred face; 7 middle-left embarrassed side pose with visible blushing face and one raised hand; 8 middle-second smiling bashful pose with visible closed eyes; 9 middle-fourth small front-facing curious pose with visible open eyes; 10 middle-right crying flustered pose with both hands raised and visible tears; 11 bottom-left hand-on-hip standing pose with blurred face; 12 bottom-second side-facing pointing pose with blurred face; 13 bottom-center small hand-on-hip pose with blurred face; 14 bottom-fourth simple standing pose with blurred face; 15 bottom-right dynamic playful pose with visible face, flowing hair, one hand making a peace sign and the other arm extended.

Visual style: Modern Japanese anime character-sheet style, glossy dark hair highlights, soft blush where visible, clean red-and-white color palette, elegant shrine maiden aesthetic, balanced proportions, subtle shadows beneath feet, no background scenery.

Constraints: No words, no logos, no watermark, no extra characters, no props. Keep all 15 figures recognizable as the same character and keep the miko outfit identical across every pose.

Goal: Create a clean, high-end Japanese B2B SaaS pitch deck / service concept board for {argument name="service name" default="NOVALITH"}, showing exactly 9 widescreen slides arranged in a 3 by 3 grid as if previewing an LP, mobile top, banner, and presentation materials.

Canvas: One landscape image containing 9 separate 16:9 slide thumbnails, evenly spaced with thin pale borders. White background, navy and teal corporate palette, light cyan diagonal line accents, subtle geometric dots, and a teal triangular corner ribbon on the upper-left of every slide containing the slide number 1 through 9 in white.

Visual style: Modern Japanese startup / enterprise SaaS design, polished vector UI, minimal gradients, soft shadows, rounded cards, thin teal data lines, navy headings, plenty of whitespace. Use a stylized geometric N logo in teal and navy, with spaced uppercase wordmark {argument name="service name" default="NOVALITH"}. No people or photos; use only icons, charts, dashboards, and abstract product visuals.

Text language: All visible slide copy should be in Japanese. Use crisp, presentation-ready typography with bold navy headlines and smaller explanatory subtext.

Slide count and layout details: Include exactly 9 slides:
1. Cover slide. Large NOVALITH logo centered above the headline 「データを、未来の意思決定に変える。」 and subheading 「中小企業のためのデータ予測プラットフォーム」. Decorative teal diagonal accents in the corners.
2. Data integration slide. Headline 「すべてのデータをひとつに統合」 with short subtext about automatically collecting and integrating scattered data. Center shows a stacked isometric database/platform cube with the N logo on top. Four connected input/output cards: 「データベース」, 「CSV / ファイル」, 「SaaS」, 「API」.
3. Forecasting slide. Headline 「高精度な予測で未来を先読み」. Large line chart card titled 「売上予測（12ヶ月）」 with legend 「実績」 and 「予測」, months along the x-axis, an upward teal trend line, and a floating KPI card reading 「予測値（来期）¥28,450,000」.
4. Risk detection slide. Headline 「リスクを事前に検知し迅速に対応」. Left chart card titled 「リスクスコアの推移」 showing a declining zigzag teal line with dates 4/1 to 4/29 and y-axis levels 高, 中, 低. Add a warning callout 「リスク上昇を検知」. Right side has exactly 3 feature rows with circular icons: anomaly detection, alert notification, and rapid response.
5. Dashboard slide. Headline 「ダッシュボードで全体を見える化」. Show a full SaaS dashboard screenshot: dark navy left sidebar with logo and exactly 6 menu items, and a main area with KPI cards, a donut chart, and a sales forecast line chart. Visible KPI examples include 「売上予測（今月）¥28,450,000」, 「需要予測（今月）15,782」, and 「リスクスコア 24/100」.
6. Automated report slide. Headline 「レポートを自動生成し共有もかんたん」. Center a white report card titled 「月次サマリーレポート」 with bullet points showing positive percentages such as +18.6% and +12.4%, a NOVALITH logo in the corner, a teal bar chart on the right, and a teal button labeled 「レポートをダウンロード」.
7. Security slide. Headline 「セキュアな環境で安心して利用」. Central shield-and-padlock illustration with floating security particles. Bottom has exactly 3 security feature icons and labels: 「暗号化通信（SSL/TLS）」, 「アクセス制御（権限分割）」, and 「定期バックアップ（自動）」.
8. Simple steps slide. Headline 「シンプルなステップですぐに始められる」. Show exactly 4 horizontally connected step cards with dotted arrows: step 1 database icon 「データを接続」, step 2 node/network icon 「モデルを選択」, step 3 play icon 「予測を実行」, step 4 bar-chart icon 「結果を確認・活用」. Each card includes a short explanatory line underneath.
9. Closing CTA slide. Headline 「さあ、データの力で未来をつくりましょう」 with subtext 「NOVALITHが、あなたのビジネスの成長をデータで支えます。」. Large centered logo and wordmark. Bottom has exactly 2 call-to-action buttons: filled teal button 「無料で試してみる　→」 and outlined button 「資料をダウンロード　↓」.

Constraints: Maintain consistent brand colors, logo, spacing, slide numbering, and diagonal teal corner motif across all 9 slides. The overall image should look like a single polished overview sheet of presentation thumbnails, not isolated full-screen slides. Avoid extra slides, extra buttons, stock photos, watermarks, or unrelated text.

Goal: Create a cute nursery-style Japanese hiragana learning poster on a light wood wall, as if generated from text only, with a slightly imperfect AI-made alphabet chart look.

Canvas: Square 1:1 image, softly lit by daylight from a window on the left. The poster is an off-white sheet pinned to the wall with four small round wooden tacks, one near each corner. Include a small potted plant and low bookshelf partially visible at the lower left for a cozy classroom or child room setting.

Main subject: A large hiragana chart poster titled {argument name="headline text" default="あいうえおひょう"} in rounded pink-red lettering with a white outline at the top center. Below the title, make a neat grid of exactly 63 rounded rectangular tiles, arranged as 9 rows by 7 columns. Each tile contains one large dark navy handwritten-style kana or punctuation mark, sometimes with tiny superscript-like numbers, dakuten marks, or small decorative icons. The text should look charming but not fully correct, like an AI-generated hiragana table.

Grid content: Use exactly 63 visible tiles. Row 1: あ, へ, う, よ, ー with tiny blue fish, う, あ with tiny apple. Row 2: と, ば, う, て, い with tiny apple, う, bunny icon. Row 3: い, ん with small 6, え, い with tiny yellow fruit, お with small 1, い with tiny apple, う with tiny orange. Row 4: お, い with comma, お with tiny green grapes, り with tiny apple, ま, す with tiny rabbit face, ば with small 6. Row 5: じ with tiny apple, い with tiny yellow chick face, さ with tiny pink fruit, い with small 2, あ with small 2, う with tiny green grapes, あ with small 2. Row 6: て with comma, よ with small 3, ち with tiny chick face, と with tiny green grapes, 々, か with tiny yellow fruit, が with tiny orange. Row 7: も with exclamation, い with tiny chick face, あ with small 2, よ, う with tiny chick face, ひ with tiny apple, も with small 2. Row 8: お with small 3, と, ら with small 3, え, ! with tiny blue fish, す with small 3, ! with tiny squirrel icon. Row 9: あ with tiny apple, う with tiny apple, て with tiny cat icon, ら with tiny orange, お with small 1, お with tiny f-like mark, ひ with tiny yellow chick.

Visual style: Soft warm pastel palette, clean rounded poster design, child-friendly educational wall chart, subtle paper texture, thin pale beige tile borders, dark blue brush-calligraphy kana, small kawaii fruit and animal stickers. Use {argument name="wall material" default="pale vertical wood paneling"}, {argument name="poster paper color" default="warm ivory"}, and {argument name="letter color" default="deep navy blue"}.

Constraints: Keep the image square, front-facing with slight natural perspective, no watermark, no extra text beyond the title and grid characters, and keep the poster readable while retaining the whimsical imperfect AI-generated kana-chart quality.

Create an ultra-realistic macro photograph from inside a transparent glass cat food bowl looking outward at a {argument name="cat type" default="brown tabby cat"} licking the food. Use an extreme wide-angle fisheye perspective with the circular glass rim forming a bright distorted ring around the frame. The cat’s face fills the upper center: pink nose close to camera, striped tabby muzzle, sharp white whiskers radiating left and right, partially visible eyes above the nose, ears near the top edge. The cat’s {argument name="tongue color" default="wet pink"} tongue is extended straight down into the bowl and dominates the center foreground, showing highly detailed rough papillae texture, moisture, tiny glossy highlights, and a realistic damp surface. The bottom third of the image is filled with {argument name="food type" default="brown dry cat kibble"}, individual pellets large in the foreground with shallow depth of field and warm golden-brown texture. Background beyond the bowl is softly blurred indoor space with warm ceiling lights, subtle plant shapes, and window reflections. Lighting is warm natural indoor daylight with specular reflections on the glass rim and tongue, crisp macro detail on the tongue and muzzle, cinematic realism, 16:9 horizontal composition, no text, no watermark.

Goal: Create a dramatic anime-style mahjong victory scene showing a young woman celebrating an enormous winning hand, with flashy Japanese game-result typography and a detailed mahjong table in the foreground.

Canvas: 4:3 horizontal illustration, high-resolution anime key visual, warm cinematic lighting, shallow depth of field, sparkling highlights, drifting pink cherry blossom petals, no watermark.

Main subject: A {argument name="character name" default="beautiful anime mahjong player"} seated behind a green automatic mahjong table. She has long flowing dark brown hair, a pink flower hair ornament with dangling ribbons, a frilly white blouse with black trim, and a jeweled brooch at the collar. Her face is intentionally obscured by a plain square censor block centered over the face. Pose: she thrusts one open hand dramatically toward the viewer while the other hand is clenched in excitement. Convey triumphant joy and surprise.

Foreground mahjong table: Show a green felt mahjong table with rails and an automatic center mechanism. Display exactly 14 front-row winning tiles, left to right: one-man, nine-man, one-man, nine-man, one-pin, nine-pin, East, South, West, North, green dragon, red dragon, plus two additional honor/terminal tiles partly matching the kokushi-musou theme if needed for spacing. Include extra side walls of face-down yellow-backed tiles and a few visible discarded tiles on the table, but keep the winning row dominant and readable.

Large victory text: On the upper left, add huge glowing gold Japanese calligraphy reading {argument name="main victory text" default="国士無双"}, painted over a black ink-splash burst with star-like sparkles. Add a smaller red-and-gold label reading {argument name="score label" default="役満"} near the lower left of the calligraphy.

Score panel: Add a dark translucent rectangular score box on the left side below the title. It contains exactly 4 player rows and a small header. Header text: 「東 4局　0本場」. Row 1: 「ロン」 and 「美少女　25000」. Row 2: 「北家　-8333」. Row 3: 「西家　-8333」. Row 4: 「南家　-8334」. Use orange for the winning call, pink for the winner name, and pale blue for negative scores.

Speech bubble: On the right side near the character, add a white rounded speech bubble with a small tail and a pink heart. Text inside: {argument name="speech bubble text" default="国士無双！\nやったぁ～～～！！"}. Use bold black Japanese lettering.

Background: Interior of an elegant mahjong room or club with warm wood tones. Include exactly 2 visible background objects: a framed vertical calligraphy sign on the wall reading {argument name="wall calligraphy" default="雀聖"}, and a gold trophy on a shelf or cabinet. Keep them slightly blurred so the table, character, and victory text remain the focus.

Visual style: glossy modern anime illustration, detailed clothing folds, luminous skin shading, dynamic hair motion, gold glow effects, ink-brush energy, celebratory game UI overlay, dramatic perspective from table height. Constraints: keep all Japanese text legible, preserve the face-covering square block, do not add extra characters, do not add logos or watermarks.

Goal: Create a black-and-white Japanese chibi drawing-song manga worksheet titled {argument name="title text" default="AIニケちゃんの かおの えかきうた"}, showing a step-by-step guide for drawing a cute simplified anime girl face.

Canvas: Tall vertical poster, approximately 2:3 aspect ratio, white background, clean monochrome line art. Use thick rounded black outlines, playful handwritten Japanese typography, and a kawaii instructional-comic style. Add small decorative sparkles, hearts, and tiny yellow accent hearts around the title.

Layout: Top title area, then an 8-panel grid arranged as 2 columns by 4 rows. Each panel has a rounded rectangle border, a black circular step number badge in the upper-left corner, and a short Japanese lyric caption across the top. The drawing inside each panel gradually becomes more complete.

Text content: Use exactly 8 numbered captions:
1. {argument name="step 1 caption" default="まるいおかおを かきましょう♪"}
2. {argument name="step 2 caption" default="ぱっつんまえがみ ふわっとね♪"}
3. {argument name="step 3 caption" default="きらきらおめめを ふたつポン♪"}
4. {argument name="step 4 caption" default="まゆげ おはな おくちで にっこりん♪"}
5. {argument name="step 5 caption" default="ほっぺのよこに けさきが たらり♪"}
6. {argument name="step 6 caption" default="あたまのうえを まるくして♪"}
7. {argument name="step 7 caption" default="うしろで きゅっと ポニーテール♪"}
8. {argument name="step 8 caption" default="AIピンで できあがり♪ AIニケちゃん♪"}

Panel details: Panel 1 shows only a large simple oval/circle face outline centered. Panel 2 adds rounded bob-like hair with straight bangs and pointed locks over the face circle. Panel 3 adds two large sparkling anime eyes with heavy upper lashes and highlights. Panel 4 shows the same head but the lower central drawing area is covered by a large pale gray vertical rectangular blur, obscuring the facial details; only the upper head outline is visible behind it. Panel 5 shows the face with bangs, big eyes, and two side hair strands falling beside the cheeks. Panel 6 adds a rounded outer head/hair outline above the face plus small eyebrows, tiny nose, and smiling mouth. Panel 7 adds a high side ponytail on the upper right with a scrunchie-like band, but the center and lower face are obscured by a large gray gradient rectangle. Panel 8 shows the near-final head with side ponytail and scrunchie, but the central face/lower area is obscured by a similar large gray gradient rectangle.

Visual style: crisp manga coloring-book line art, no color fills except black outlines and minimal yellow heart accents near the title; thick friendly strokes, rounded corners, cute educational worksheet look. Keep the drawings simple, deformed, and childlike, as if for a drawing-song tutorial.

Constraints: Use exactly 8 panels and exactly one step number per panel. Preserve the Japanese text as written. Do not add extra panels, extra characters, background scenery, watermarks, or realistic shading.

Create a horizontal WordPress plugin banner with a refined minimalist Japanese aesthetic. Canvas: wide rectangle, approximately 1540 × 500 pixels, pale warm peach background with a very subtle soft radial glow near the center. Add a thin medium-brown rectangular border inset slightly from the edges. Layout: place a large stylized folding fan icon on the left occupying about 35% of the width, and place the product name and tagline on the right, vertically centered. Icon details: a simple geometric Japanese folding fan in dark warm brown outlines and fills, angled slightly upward, with exactly 7 visible fan panels separated by thick brown ribs, cream-filled inner panels, dark brown filled outer side sections, and a rounded dark brown pivot circle at the bottom center of the fan. Text: large serif title reading {argument name="plugin name" default="Noshi-Kanamer"} in dark warm brown, bold, elegant, with generous letter spacing; directly underneath, smaller sans-serif tagline reading {argument name="tagline" default="The WordPress developer toolkit for clean handoffs."} in the same brown. Keep the design clean, flat, vector-like, professional, and suitable for a WordPress plugin banner. Use no extra icons, no shadows, no mockups, no people, and no additional text.

Using the provided reference image as the product concept and visual style, redesign it as a vertical gachapon-style promotional poster for miniature torso figures. Keep the cute jellyfish-themed fashion-doll merchandise feeling, but change the composition into a soft lavender-and-cream catalog layout with a dreamy bedroom/cosmetic-shelf background, pearl and lace decorations, and three torso figures displayed across the center on dark wooden mannequin stands.

Create exactly 3 miniature torso figures, each with a numbered name label above it:
1. 「01 ゆめみるくらげ」 — a pale white-and-light-blue fluffy dress based on the existing soft white jellyfish outfit from the reference.
2. 「02 くりすますくらげ」 — a new red Christmas-themed jellyfish dress with white fur trim, capelet, black belt, and red-and-white ribbon tentacles.
3. 「03 ほしぞらくらげ」 — a new lavender-and-purple starry jellyfish dress with translucent sleeves, gold decorative chains, layered skirt, and long pale purple ribbon tentacles.

Top header: Use the brand text {argument name="brand name" default="くらげ堂"}, the main title {argument name="headline text" default="くらげちゃんの ミニトルソーフィギュア"}, and the subtitle {argument name="subtitle text" default="～全３種～"}. Add a small boxed notice in the upper right with Japanese capsule-toy information: 「対象年齢 15才以上」「1カプセル 1個入り」. Add a small pearl-framed badge reading 「全3種」.

Bottom section: Add a wide lavender scalloped information band. On the left, show the size note 「サイズ 全高 約6.5cm」 with a tiny standing mannequin icon and vertical arrow. In the center, place a lace-framed text box explaining the product in Japanese: 「くらげちゃんの３つの衣装が かわいいミニトルソーフィギュアに♪ お部屋に飾って楽しめます。」 followed by three small notes: 「※画像はイメージです。」「※台座は取り外しできません。」「※自立します。」 On the right, add an ornate circular price badge with {argument name="price text" default="全３種 400円（税込）"}.

Style constraints: photorealistic miniature collectible figures, delicate translucent fabrics, soft pastel lighting, polished catalog photography, high-end Japanese capsule-toy flyer design. Preserve the torso-figure concept and jellyfish fashion motif from the reference, but use this new vertical poster layout and the three specified outfit variants. Do not add extra figures, extra outfit labels, watermarks, or unrelated text.

Create a wide 16:9 anime-style illustration of one {argument name="character name" default="faceless white-haired schoolgirl pitcher"} standing alone at the exact center of a sunny baseball field, viewed from the front at full body distance. She has short pale silver-white bobbed hair with straight bangs, a deliberately blurred or featureless face, and a slim teenage build. She wears a white long-sleeve blouse under a dark charcoal pleated pinafore-style school uniform dress with a belt, white knee socks, and white Mary Jane shoes, and she holds one black baseball glove with both hands in front of her chest as if ready to pitch. The setting is an empty school baseball diamond: brown dirt infield filling the foreground, a circular pitcher’s mound under her feet, clean white foul lines and home-plate markings behind her, green outfield grass beyond, a tall central backstop fence with netting, low green outfield walls, two tall light poles at the far left and far right edges, and dense leafy trees in the background. The sky is vivid bright blue with scattered fluffy white clouds, giving a clear summer afternoon feeling. Use crisp Japanese anime background art, clean linework, vibrant colors, soft sunlight, balanced symmetrical composition, and high detail in the field textures and fence. Keep the camera level and centered, no text, no watermark, no extra characters, no baseball bat, no motion blur.

Create a polished anime-style illustration of {argument name="character names" default="Yuri and Liangfang"} sparring inside a traditional martial arts dojo. The scene contains exactly 2 barefoot young women with fox/cat-like animal ears, both in dynamic fighting stances on green tatami mats: the left character has short brown hair, brown animal ears, a navy cropped jacket over a black fitted training outfit with black shorts and wrist guards, stepping wide with one palm raised defensively; the right character has short black hair, black animal ears, a tan blazer over a black fitted training outfit with black shorts, throwing a straight punch toward the left character while guarding with her other fist. Both faces should be softly blurred or indistinct, with no clear facial features. Add a magical soap-bubble energy effect between their hands: exactly 3 large iridescent transparent bubbles or bubble-rings, plus exactly 10 smaller floating bubbles, all catching rainbow highlights and forming a swirling arc around the punch. The dojo has wooden beams, sliding-style walls, a wall mirror on the right reflecting gym equipment, shelves with blue exercise balls and training gear, framed pictures, warm sunlight, and a large window on the left showing snowy mountains and pine trees outside. Use cinematic composition, soft natural lighting, detailed anime rendering, clean linework, realistic perspective, glossy highlights on skin and bubbles, no text, no watermark.

Create a polished 3D cartoon editorial scene in a cozy modern interview lounge, formatted as a wide 16:9 image. Two anthropomorphic AI-logo mascot characters sit facing each other in beige upholstered armchairs around a round wooden coffee table. On the left, a black rounded-square mascot body with a large white OpenAI knot logo on its front has black tube arms and legs, white gloved hands, and oversized white sneakers; it is gesturing with one raised index finger as if making a point. On the right, a tan rounded-square mascot body with large black letters “AI” on its front sits with crossed legs and black gloves/boots, leaning back skeptically. Add exactly 2 speech bubbles: the left large bubble says “i’m not charging twice so that developers feel respected instead of exploited”, with “respected” colored green and “exploited” colored red; the right smaller bubble says “that seems difficult to monetize”. Include exactly 3 visible text signs: a small lit shelf sign on the left reading “BUILD SHIP REPEAT”, a framed wall poster on the right reading “ANTHROPIC / SAFETY FIRST* / *TERMS APPLY”, and the black mug logo mark on the table. The coffee table should contain exactly 2 objects: a black mug with a white OpenAI-style logo and a small potted succulent. Background details include dark charcoal walls, warm studio lighting, a bookshelf with plants on the left, a large potted plant on the right, a soft round rug, and shallow depth of field. Use a satirical tech-industry tone, premium toy-like 3D rendering, soft global illumination, expressive body language, clean readable lettering, and no extra characters or additional text. The left speech bubble text may be customized as {argument name="left speech bubble" default="i’m not charging twice so that developers feel respected instead of exploited"}; the right speech bubble text may be customized as {argument name="right speech bubble" default="that seems difficult to monetize"}; the wall poster headline may be customized as {argument name="wall poster text" default="ANTHROPIC / SAFETY FIRST* / *TERMS APPLY"}; the left character logo may be customized as {argument name="left mascot logo" default="OpenAI knot logo"}; the right character label may be customized as {argument name="right mascot label" default="AI"}.

Create a dramatic vertical anime action illustration of {argument name="character description" default="a fierce young female martial artist with very long flowing pink hair and green eyes"} inside {argument name="setting" default="a traditional Japanese wooden dojo with tatami floors, shoji windows, rafters, and warm dusty sunlight"}. She is lunging forward low to the ground in a powerful spinning kick, wearing {argument name="outfit" default="a loose white karate gi with a black belt, wrapped fists, and a slightly open collar"}; her hair streams wildly across the frame in long ribbon-like strands. On the right side, a large rectangular straw tatami target mat is tilted upright at a steep angle, its edge bursting with splinters and straw debris from impact. Show exactly ten visible metal shuriken: eight embedded in the upright tatami target at different heights and angles, and two still flying through the air near the character. Emphasize speed with circular motion blur around her extended kicking leg, dust clouds, flying straw fragments, sparks, and dynamic diagonal composition. Use {argument name="visual style" default="high-detail cinematic anime, semi-realistic rendering, sharp linework, volumetric lighting, shallow depth of field"}. The mood should feel like {argument name="mood" default="an intense ninja training strike captured at the exact moment of impact"}. No text, no watermark, no extra characters.

Create a grainy cinematic 35mm street-photography style image inside an old European café at night, with a muted green-yellow color cast, soft window reflections, low available light, shallow depth of field, and the feeling of a candid documentary photo. The scene shows exactly two seated patrons: on the left, {argument name="left patron" default="a woman with long dark hair in a black coat, holding a cigarette near her face"}, and on the right, {argument name="right patron" default="a man in a gray coat sitting behind her at another table, looking downward at a newspaper"}. Both faces are intentionally obscured by plain dark square blocks for anonymity. Include exactly three round marble café tables: one foreground table on the left with a white espresso cup and saucer, a small drinking glass, a black ashtray, and a cigarette pack; one middle-right table with a white espresso cup and saucer, a clear water glass, and folded newspapers; and one blurred empty foreground table at the bottom right. Include exactly three visible bentwood café chairs: one behind the middle-right table, one partially visible at bottom right, and one dark chair shape near the left patron. Add an old wall menu board in the upper left with hand-lettered text reading “PORTO ROUGE FINS C 250.” The background should contain dark wood paneling, mirrors or glass doors, a vague reflected figure, and a bright greenish doorway or window on the far right. Use {argument name="lighting mood" default="dim smoky café lighting with soft highlights and deep shadows"}, {argument name="film stock look" default="aged color negative film, visible grain, slight blur, vignetting"}, and {argument name="aspect ratio" default="4:3 horizontal frame"}. Avoid modern objects, crisp digital sharpness, clean studio lighting, or any additional people in the foreground.

Create a cinematic 35mm film photograph of a lone person in a long dark coat standing near the center of an abandoned industrial warehouse at golden hour. The building is vast, decayed, and partially ruined, with tall broken window frames across the back wall, exposed steel beams, cracked concrete columns, peeling paint, rusted railings, debris, dry weeds, and overgrown plants pushing through the floor. The low sun shines directly through the left-center windows, producing a bright warm flare and long dramatic shadows stretching across the dusty concrete toward the camera. Compose in a wide 16:9 horizontal frame with the subject small against the architecture, slightly right of center, facing the glowing windows, creating a quiet contemplative mood. Use natural warm amber sunlight mixed with deep cool shadows, high dynamic range, realistic texture, visible film grain, slight softness, and analog color grading. Add strong accidental film light leaks: a vertical saturated red-pink band along the far left edge and a warm orange-red flare bleeding in from the right edge. Photorealistic, atmospheric, documentary-style urban exploration photography, no text, no watermark.

Create a cinematic black-and-white fine-art photograph of {argument name="subject" default="a ghostly ballerina"} dancing alone inside {argument name="location" default="an abandoned old theater or ballroom"}. The dancer is positioned slightly right of center, wearing a translucent knee-length ballet dress and pointe shoes, with her arms extended in an elegant, fragile pose; use a long-exposure multiple-exposure effect so her body appears as two or three overlapping semi-transparent silhouettes, suggesting motion and haunting presence. The room is dark, decayed, and atmospheric, with cracked plaster, an arched alcove or stage opening on the right, rows of shadowed seats or debris in the background, and tall broken arched windows on the left. Strong sunlight blasts through the window from the left rear, creating visible shafts of dusty light, smoky haze, high-contrast rim lighting, and a bright rectangular patch of light on the wooden floor beneath the dancer. Use {argument name="lighting style" default="dramatic backlight with volumetric dust beams"}, deep blacks, glowing highlights, film grain, soft blur from movement, and a moody gothic mood. Compose in a horizontal 4:3 frame, wide interior view, low-to-mid camera height, with the dancer small enough to show the environment. No text, no modern objects, no color, no clean studio look.