GPT Image 2 prompts

Using the provided reference image, transform the person into an Edo-period ukiyo-e / nishiki-e / kabuki actor portrait style while preserving the original pose, black bucket hat, hand-near-mouth gesture, and anonymous blurred face area. Replace the modern shirt with traditional layered kimono robes: a dark navy outer robe with two circular floral crests, a tan checked inner robe, and subtle reddish trim. Change the composition to a vertical close-up woodblock print on aged beige washi paper with visible grain, ink texture, worn edges, and muted traditional pigments. Add exactly 4 new decorative/background elements: 1 vertical title cartouche on the upper left with Japanese calligraphy reading {argument name="vertical title text" default="東都新賢似顔繪"}, 1 red artist seal beneath it reading {argument name="seal text" default="豊國画"}, and 2 stylized blue-gray Edo cloud motifs, one near the upper right and one along the left middle. Keep the face intentionally obscured with a soft rectangular blur matching the reference privacy treatment. Avoid photorealism, modern lighting, and extra text.

Goal: Create a vertical premium coffee shop grand-opening promotional poster for {argument name="brand name" default="牛顿咖啡"}, with a warm beige marble-paper background, brown typography, taped paper collage elements, and a realistic mall storefront mockup in the center.

Canvas: Portrait poster, 2:3 ratio, high resolution. Use an elegant cream, caramel, coffee-brown, and warm gold palette. The overall mood is modern Chinese commercial advertising, cozy, polished, and upscale.

Top section: Center a simple coffee-cup logo with steam above the brand name {argument name="brand name" default="牛顿咖啡"} and small English text “NEWTON COFFEE”. Below it place spaced uppercase text “NEW STORE OPENS”. The main oversized headline should read {argument name="headline text" default="拍围挡 · 喝咖啡"} in bold dark-brown Chinese serif display type.

Coupon banner: Under the headline, add a torn white paper strip taped at the upper-right corner with beige masking tape. On the left of the strip draw a small line-art steaming cup and handwritten “Enjoy!”. In the center place a huge “1000”. On the right place Chinese and English offer text: “杯免费送”, “FREE COFFEE”, “FREE COFFEE”.

Central storefront scene: Create a realistic curved mall construction hoarding / storefront wall with glossy floor reflections and glass mall architecture behind it. The hoarding is wrapped in coffee-themed graphics with swirling latte foam and marble textures. It contains exactly 5 visible collage areas: 1) left vertical torn-edge coffee photo showing latte art in a cup, 2) small taped left notice reading “新店开业 / 1000杯 / 免费送 / FREE COFFEE”, 3) large central cream panel with the logo, brand name, and handwritten slogan {argument name="slogan" default="好咖啡，这里喝得到！"}, 4) right taped notice reading “NEW STORE OPENS / 2026 / 6/9 / 我们在这里等你”, 5) lower-right torn photo of a warm coffee shop interior with a small cup illustration beside it.

Central feature icons: Along the bottom of the large central panel, show exactly 3 minimal brown line icons with labels: 1) coffee bean icon labeled “现磨咖啡 / FRESHLY GROUND”, 2) milk carton icon labeled “鲜奶拿铁 / FRESH MILK LATTE”, 3) dessert cloche icon labeled “手作甜点 / HANDMADE DESSERT”.

Instruction section below storefront: On a clean cream background, create exactly 3 horizontal participation steps separated by arrows: 1) camera icon with “拍围挡打卡”, 2) cup icon with “领免费咖啡”, 3) smiley icon with “进店体验”. Beneath the steps, add two centered lines of smaller Chinese instruction text: “拍下围挡任意角度照片，发布至社交平台并带话题 #牛顿咖啡新店开业#” and “即可到店领取免费咖啡一杯！每日限量，先到先得！”

Bottom information row: Create exactly 3 information blocks with thin vertical dividers: 1) location pin icon, “牛顿商场名”, “NEWTON MALL”; 2) stacked-floor icon, “L1层”, “LEVEL L1”; 3) calendar icon, {argument name="opening date" default="2026 - 6 - 9"}, “盛大开业”. At the very bottom add handwritten Chinese text “牛顿喝了都说好的咖啡” with a casual underline.

Visual style: Elegant Chinese coffee retail poster, photorealistic storefront mockup combined with paper collage, torn paper edges, masking tape, soft shadows, warm lighting, subtle marble texture, coffee foam swirls, refined typography, balanced central alignment.

Constraints: Keep all listed text legible and in the same approximate positions. Use exactly 3 feature icons, exactly 3 participation steps, exactly 3 bottom information blocks, and exactly 5 visible collage areas on the storefront. No extra logos, no watermark, no QR code.

Create a vertical 9:16 ultra-detailed fantasy fashion portrait of an anime-inspired young woman with {argument name="hair color" default="cotton-candy pink"} hair styled into exactly 2 long flowing twin tails, drifting dramatically left and right as if underwater or in zero gravity. Her face is intentionally hidden by a flat centered square censor block in muted dusty pink, covering the entire facial area. She wears a black gothic couture mini dress with sheer translucent sleeves, lace mesh, sparkling beadwork, and abundant pink floral appliqués. Include exactly 7 large prominent pink roses on the outfit and nearby foreground: one on the chest, three clustered at the waist, one lower on the skirt, one at the left edge, and one soft blurred rose in the lower right foreground. Add black-and-pink floral hair ornaments on both sides of her head, a black lace choker, delicate sheer arm ruffles, and tiny white bead strands. Surround her with a dreamy surreal background of floating translucent bubbles, glossy black pearls, pink pearls, black-and-pink patterned spheres, circular vinyl-record-like discs, and fine botanical line-art flowers. Make exactly 3 prominent circular disc motifs visible: one large at the upper right, one small at the lower left, and one partial at the bottom right. Make exactly 10 prominent floating glossy spheres visible in the foreground and midground, mixing pink, black, clear, and patterned surfaces, with additional tiny sparkle particles kept subtle. Use a luminous white and pale pink background, high-key lighting, shallow depth of field, soft bokeh, glossy reflections, delicate black ink filigree, and a luxurious hyperreal digital-art finish. Color palette: {argument name="accent palette" default="pastel pink, black, white, and pearl silver"}. Pose: upper body angled slightly left, one bare shoulder visible, right hand raised gracefully near the face, elegant and ethereal. No text, no watermark, no extra characters.

Create a cinematic anime-style illustration of two chibi cat-eared motorcycle girls riding side by side along a waterfront road at sunset, viewed from behind and slightly to the left as they travel toward the glowing horizon. The scene shows exactly 2 riders: the left rider has long flowing {argument name="left rider hair color" default="lavender purple"} hair, a glossy pale helmet with cat ears, a black riding suit and gloves, and rides a green mini sport bike labeled Kawasaki and KSR110 with a license plate reading KSR110; the right rider has wavy {argument name="right rider hair color" default="light brown"} hair, a glossy pink cat-ear helmet, a brown jacket, dark pants, tall boots, and rides a small black Honda-style bike with a license plate reading CT125. Both riders are cute compact chibi proportions but rendered with detailed mechanical realism on the motorcycles, including visible wheels, suspension, mirrors, exhausts, taillights, brake discs, chains, and handlebars. The background is a detailed Tokyo Bay waterfront with calm reflective water, the Rainbow Bridge stretching across the right half of the image, the Fuji TV building and city high-rises on the left, guardrails and lane markings in the foreground, and exactly 3 seagulls flying in the orange-purple sky. Use a warm golden-hour palette with the sun low near the bridge, shimmering reflections on the water, long shadows on the asphalt, soft clouds tinted orange and pink, and crisp high-detail anime rendering with painterly lighting. Make the image horizontal 4:3, immersive, polished, and free of captions or watermarks, while preserving the visible bike text and plates: {argument name="left bike plate" default="KSR110"} and {argument name="right bike plate" default="CT125"}.

Create a dramatic vertical high-fashion portrait of a young Japanese woman in an ornate black kimono with vivid red accents and white cherry blossom patterns, seated in a dynamic swirling pose. Her face is intentionally covered by a centered soft-edged rectangular blur in warm beige-brown tones, while long dark hair flows sideways through elaborate avant-garde hair ornaments made of red and black lace, cords, netting, and floral pieces. The kimono should have wide flowing sleeves, a black obi belt, layered red lining, and one visible bare thigh through the fabric folds. Surround her with an abstract explosion of Japanese ink-art elements: red paint splatters, black calligraphic brush arcs, tangled circular linework, delicate mesh nets, cherry blossom branches, red flowers, and fine radial thread-like lines, all wrapping around her body as if caught in motion. Use the color palette {argument name="color palette" default="deep black, crimson red, white blossoms, pale skin, and off-white background"}. Make the composition {argument name="composition" default="vertical 9:16, tightly cropped, subject angled diagonally from upper left to lower right"}. Use {argument name="visual style" default="hyper-detailed editorial fashion photography mixed with surreal Japanese ink illustration"}, with sharp fabric textures, glossy black hair, intricate lace detail, shallow depth of field in the foreground flowers, and a clean pale studio background. The mood should feel {argument name="mood" default="elegant, chaotic, gothic, and ceremonial"}. No readable text, no watermark, no extra people.

Create a cinematic photorealistic blog cover image showing a chaotic crowd of zombie-like shoppers pressed against the glass entrance of a Japanese retail store on a bright sunny day. The camera is inside the store looking outward through automatic sliding glass doors, wide 16:9 composition, low slightly wide-angle perspective, intense backlight, blue sky and small clouds outside, reflections on the glass, metal door frames, posters on the side windows, and a stack of red shopping baskets at the lower right. About 14 pale, dirty, frantic people in torn light-colored clothes crowd the doorway with outstretched hands and smeared grime, like a consumer stampede. Exactly 10 visible smartphone screens are held up toward the camera, each displaying large red Japanese sale text {argument name="phone screen text" default="本日ポイントアップ↑"} on a white background with small red confetti dots; include 3 large foreground phones, 5 raised background phones, and 2 smaller partially visible side phones. Include 1 large raised rectangular placard near the upper left-center with the same red Japanese text, and 1 bright red-and-yellow store poster on the right window advertising {argument name="poster sale text" default="ポイント 5倍"}. Add exactly 9 opaque dark gray rectangular anonymization blocks covering faces in different sizes, including one large block in the center foreground, one large block on the right foreground, one large block on the left foreground, and six smaller blocks across the background crowd. Keep the scene realistic rather than cartoonish: gritty clothing texture, bluish corpse-like hands, motion blur on reaching arms, shallow depth of field, high-contrast daylight, documentary disaster-movie atmosphere, no extra readable English text, no watermark.

Create a high-energy square anime illustration of {argument name="character name" default="Becky"}, a cute blonde cowgirl, taking an extreme close-up selfie with {argument name="donkey name" default="Bourbon"}, a goofy gray donkey. The girl is on the left foreground with huge expressive blue eyes, one eye winking, flushed cheeks, a panicked funny open mouth with her tongue sticking out, sweat drops, and messy curly golden-blonde hair spilling from under an oversized brown leather cowboy hat with stitched edges and silver conchos. She wears a brown leather western jacket, red paisley bandana, cream lace-trimmed top, denim shorts, belt with a western buckle, and cowgirl details. The donkey is on the right, pushing into frame with tall ears, wild eyebrows, big mismatched-looking surprised eyes, a broad toothy grin, and its mouth biting or tugging on the fringe of the girl's leather jacket, making the pose chaotic and comedic. Use a fisheye wide-angle perspective, exaggerated chibi-anime facial expressions, glossy cel shading, highly detailed hair and leather textures, dynamic cropped composition, vibrant colors, and a psychedelic swirling background of neon pink, teal, orange, purple, and blue spiral patterns. Make it playful, loud, cute, and absurd, like a polished modern Japanese anime key visual; no text, no watermark.

Goal: Create a monochrome manga page in a delicate, glittering shoujo style, showing a shy young singer alone in a luminous rehearsal room or studio, with a sparkling, emotional atmosphere and cinematic paneling.

Canvas: Vertical manga page, black-and-white grayscale, approximately 3:4 aspect ratio, thin black outer border, crisp black panel gutters, high-detail screentone shading, soft bloom, countless star-like sparkles and dust motes across every panel.

Layout: Use exactly 4 panels. Panel 1 is a wide horizontal establishing shot across the full top third of the page. Panel 2 is the middle-left panel, a close bust shot. Panel 3 is the bottom-left panel, a close-up of hands gripping a microphone. Panel 4 is a tall right-side panel spanning the middle and bottom rows, showing the emotional main character with a speech bubble.

Subject details: The main character is {argument name="character name" default="a shy teenage girl singer"}, with {argument name="hair style" default="short messy bob hair with soft bangs"}, a slim build, downcast eyes, wet-looking lashes, faint blush, and a fragile apologetic expression. She wears an oversized light T-shirt with a visible dark camisole or tank strap at one shoulder. She holds a handheld microphone with both hands, close to her mouth, as if singing or apologizing. Keep her posture slightly hunched and inward, conveying embarrassment and vulnerability.

Panel details:
- Panel 1: Wide high-angle view of a spacious mirrored dance studio or glass-walled practice room. Reflective floor tiles, bright windows or mirrors on the walls, faint silhouettes and reflections in the background. The girl stands alone near the center, very small in frame, holding the microphone. Add musical text near her: {argument name="humming text" default="・・・・・・・♪"}.
- Panel 2: Middle-left close-up of the girl from chest up, holding the microphone near her mouth. Her face is mostly covered by a plain opaque gray rectangular censor block centered over the face. Add small music/sound text on the left: {argument name="soft sound effect" default="♪…っ"}. The background is filled with bright sparkles.
- Panel 3: Bottom-left close-up of both hands tightly squeezing the microphone handle. Fingers are slender, tense, and carefully drawn. Add Japanese sound effect text at the left: 「ギュッ」. Strong sparkles overlay the shirt and hands.
- Panel 4: Tall right panel close-up of the girl from head to waist, face visible and tilted down with teary, apologetic eyes. She clutches the microphone with both hands. Place a vertical oval speech bubble in the upper right containing the Japanese text {argument name="speech bubble text" default="あ、また…ごめんなさい！"}. Add a small MOOTOON watermark/logo in the bottom-right corner.

Visual style: High-quality grayscale manga illustration, elegant thin line art, luminous shoujo sparkle effects, soft foggy highlights, reflective glass and floor surfaces, delicate hair strands, emotional close-ups, dramatic contrast between wide isolation and intimate hand/face panels. Keep the image monochrome only, with no color except grayscale tones.

Constraints: Exactly 4 manga panels; exactly 1 speech bubble; exactly 3 visible sound/music text elements outside the speech bubble: 「・・・・・・・♪」, 「♪…っ」, and 「ギュッ」. Include the gray rectangular face block only in the middle-left panel. Avoid extra characters, extra speech bubbles, or additional readable text beyond the specified elements and the MOOTOON watermark.

Create a dramatic vertical anime-style portrait of {argument name="character name" default="a mysterious white-haired swordsman"} standing centered in a narrow futuristic city alley at night during heavy rain. He has shaggy layered {argument name="hair color" default="silver-white hair with vivid red tips near the ears"}, a confident powerful posture, and his face is deliberately covered by a plain semi-opaque rectangular censor block in muted gray-pink, placed over the entire facial area. He wears a glossy wet {argument name="coat color" default="crimson red"} long leather trench coat with large lapels, gold buttons and shoulder details, over a black open-collar shirt, black shiny pants, and a black belt with a brass buckle. His arms are crossed tightly across his chest, sleeves rolled slightly, with rain droplets and highlights on his skin and clothing. Behind him, bright electric-blue lightning cracks across the dark sky and forms jagged energy arcs around his silhouette, creating a glowing halo effect; include visible steam or mist puffs near his shoulders. The background is a tall, dark urban alley with shadowed building walls on both sides, small glowing signs and streetlights near the lower edges, deep perspective, wet reflective surfaces, and dense vertical rain streaks. Use high-contrast cinematic lighting, saturated blues and reds, glossy reflections, sharp anime illustration detail, dynamic composition, and a heroic low-angle feel. No text, no watermark, no extra characters.

Create a photorealistic vertical underwater fantasy portrait of a young woman wearing a vintage brass diving helmet with a clear spherical glass dome, seated in an ancient submerged ruin. Her face should be intentionally obscured by a centered solid square privacy block in muted brown, covering the facial features while leaving the hair, helmet, neck, and outfit visible. She has medium-length dark brown hair visible inside the dome and wears a vivid {argument name="suit color" default="red"} tailored blazer and matching trousers over a white shirt, with dark shoulder straps and brass diving gear around the collar. She holds a black handheld game console horizontally with both hands as if excitedly playing underwater. The scene is bright, cinematic, and highly detailed: turquoise-blue ocean water, sunbeams from above, rising air bubbles, colorful coral reefs on both sides, small tropical fish, two orange clownfish in the lower left foreground, a translucent jellyfish in the mid-right background, an old treasure chest on the seafloor at left, and ancient stone columns and ruins in the rear left. Use realistic lighting, glossy reflections on the glass helmet, brass metal details, soft depth of field, vibrant reef colors, and a magical adventurous mood. Vertical 3:4 composition, centered subject from knees upward, no text, no watermark.

Goal: Create a cinematic social media video thumbnail / award post image showing a fantasy shogi match transforming into a battlefield, with a dark PixVerse-style video-card presentation.

Canvas: 600×483 px landscape card on a black background with rounded corners. The upper 82% is a cinematic image still; the lower strip is black with a large title. Use warm firelight, deep shadows, shallow depth of field, dramatic movie lighting, and high-detail photorealistic animal characters.

Main scene: Inside a dim traditional Japanese room, two anthropomorphic dogs sit across a wooden shogi board that has become a miniature battlefield. On the left is {argument name="left character" default="PONTA, a pink Shih Tzu"}, small and fluffy, wearing a pale pink kimono, holding a thin wooden pointer toward the board with a confident strategic expression. On the right is {argument name="right character" default="a worried bulldog opponent"}, large and wrinkled, wearing a dark robe, gripping both sides of his head in despair. Behind them are blurred lanterns, flames, and traditional wall art.

Board details: The shogi board fills the lower center of the image in perspective. Show a dense miniature battle scene on top of the grid: exactly two opposing armies, one accented in red/orange and one accented in blue. Include about 64 visible miniature soldier figures total, mixed with shogi pieces, banners, spears, and glowing embers. A bright fiery clash burns near the center of the board. Make the wooden grid and at least one shogi piece with Japanese characters visible near the front edge.

Overlay text and UI: Add a rounded dark-brown pill label at the top left reading {argument name="award label" default="Top 10 Popular Picks"} in white sans-serif text. At the lower left of the image still, place a circular colorful avatar icon followed by white text reading {argument name="creator line" default="PONTA (Pink Shih Tzu)"}. At the lower right of the image still, place the white duration text {argument name="duration" default="00:08"}. In the black bottom area, add the large white title text {argument name="video title" default="Shogi on the Battlefield"} aligned left.

Visible discrete text elements: exactly 4 text elements: the award label, the creator line, the duration, and the bottom title. Visible major characters: exactly 2 dogs. Visible UI elements: exactly 1 rounded award pill, exactly 1 circular avatar, exactly 1 video duration marker, exactly 1 bottom title area.

Style constraints: photorealistic yet fantastical, cinematic, warm orange highlights, dark vignette, high contrast, crisp central details, no extra logos, no watermark, no additional captions beyond the specified text.

Create an epic vertical fantasy illustration of a radiant {argument name="character name" default="fiery angel"} descending from the heavens, centered in the frame and shown full body, floating above a vast mountainous kingdom at sunset. The angel wears ornate polished {argument name="armor color" default="gold and ivory"} celestial plate armor with intricate filigree, layered pauldrons, engraved chest pieces, glowing orange seams, and long flowing tattered golden cloth panels. Give the figure long wavy {argument name="hair color" default="auburn"} hair, with the face intentionally hidden by a simple flat {argument name="face covering" default="brown rectangular mask"} so no facial features are visible. Add exactly two enormous flaming feathered wings, one on each side, spread wide and arcing upward, made of bright orange, gold, and white fire with visible feather structure and sparks. Behind the head place one glowing circular sacred halo sigil with fine geometric rings and radiant runes, burning like a sun. In the angel’s left hand place one blazing {argument name="weapon" default="flaming sword"} angled down toward the lower left, with a white-hot blade and golden hilt. Surround the figure with swirling ribbons of fire forming a loose circular vortex around the body and across the lower foreground. The background is a dramatic heavenly sky with storm clouds, sunbeams, embers, distant jagged mountains, glowing rivers or lava-like light paths through the valley, and tiny castle spires on cliffs. Use ultra-detailed high-fantasy digital painting style, cinematic lighting, volumetric rays, extreme glow, sharp armor detail, warm orange-gold palette, majestic divine mood, symmetrical heroic composition, no text, no watermark.

Goal: Create a single-page Japanese manga-style instructional comic about using a 3×3 image grid with GPT and seedance to generate advertising videos, narrated by a lazy but wise capybara. Make it look like a polished social media explainer comic with warm lighting, detailed animal rendering, and hand-drawn panel borders.

Canvas: Vertical 4:5 page, about 768×960 px, cream paper background with thin black outer border. Use a warm sepia-brown palette, soft shadows, cinematic illustration, semi-realistic capybara fur, manga speech bubbles, and clean readable Japanese typography.

Main character: A large chubby capybara with tan-brown fur, sleepy half-closed eyes, relaxed smug expression, and small dark paws. The capybara acts like a laid-back advertising strategist. Optional name: {argument name="character name" default="lazy capybara strategist"}.

Layout: Use exactly 6 comic areas: 1 large hero panel at the top, 4 smaller instructional panels in a 2×2 grid underneath, and 1 wide closing panel at the bottom.

Hero panel: The capybara sits at a wooden desk in a dim cozy office, one paw under its chin, in front of an open silver laptop. On the desk include exactly 4 visible desk items: a white mug labeled 「やる気ゼロ。」, a black smartphone, a spiral notebook labeled 「今日もやらない」 with a tiny capybara doodle, and a brown leather satchel at the right. Add a dramatic white spiky manga burst on the left with bold text: {argument name="burst quote" default="これ、地味にすごいやつやん…！"}. On the right side of the hero panel place a pinned beige reference poster titled 「引用元ポスト(要約)」 and main headline 「GPT5.5 × seedance 架空広告の作り方」. The poster contains exactly 9 thumbnail images in a 3×3 grid: red sports car, sneaker, perfume bottle, burger, luxury wristwatch, resort pool, coffee cup, smartphone on desk, tropical suitcase/beach scene. Under the grid, show four short numbered Japanese steps explaining: create images in a 3×3 grid, feed them to GPT to convert into a prompt, combine images plus prompt to generate video, and create variations with #higgsfield and MCP.

Instruction panel 1: Top-left small panel with blue circle number 「1」 and title 「何が便利か」. Show capybara from the chest up, thoughtful and unimpressed. Include one speech bubble saying 「3×3をまとめてGPTに食わせるから、一貫性のあるプロンプトがサクッとできる。」 Add a small visual diagram at the bottom: a tiny 3×3 color grid arrow to a document icon labeled 「プロンプト(設計図)」. Add small caption 「1枚ずつ考えるより手戻り減るで。」

Instruction panel 2: Top-right small panel with blue circle number 「2」 and title 「初心者が真似するなら」. Capybara raises one finger like giving advice. Include a speech bubble: 「まずは3×3で好きなジャンルの画像を集めてみる。無料素材や自作でOK。」 On the right show exactly 3 vertically stacked icons with arrows: a 3×3 grid labeled 「集める」, a brain labeled 「プロンプト化(GPT)」, and a play button labeled 「動画生成(seedance)」.

Instruction panel 3: Bottom-left small panel with blue circle number 「3」 and title 「注意点」. Capybara holds up two fingers with a serious sleepy expression. Speech bubble: 「権利・商用利用はちゃんと確認やで。素材の利用規約と、生成物の使いみちは自己責任。」 Along the bottom include exactly 3 warning icons with labels: yellow warning triangle 「権利確認」, document 「規約チェック」, lock 「自己責任」.

Instruction panel 4: Bottom-right small panel with blue circle number 「4」 and title 「小さく試す方法」. Capybara leans on the panel edge, explaining calmly. Speech bubble: 「いきなり全額を狙わなくてええ。3×3→1本だけ出してみる。それを見て微調整。小さく回して学ぶのがいちばん早い。」 On the right draw a circular workflow with exactly 4 steps and arrows: 「作る(3×3)」, 「出す(1本)」, 「見る(反応)」, 「調整(改善)」.

Closing panel: Wide bottom panel. Capybara lies sprawled on the desk, eyes closed, extremely lazy, with tiny snore text 「ふぅ…」. Large text on the left: {argument name="closing line" default="今日もやらないけど、ここだけ試すならアリ。"}. Add a small thought bubble: 「動いたもん勝ち…ワイは昼寝や。」 On the right place a taped yellow sticky-note checklist titled 「まとめ」 with exactly 4 unchecked items: 「3×3で設計図を作る」, 「自動で量産できる」, 「小さく回して学ぶ」, 「今日もやらない(大事)」, plus a paw-print mark.

Visual style: Detailed semi-realistic capybara, manga/comic instructional layout, warm office lighting, soft beige shadows, expressive Japanese lettering, blue numbered badges, hand-drawn ink outlines, slight paper texture, humorous but professional social media explainer aesthetic.

Constraints: Keep all Japanese text legible and in the locations described. Use exactly 6 panels, exactly 9 poster thumbnails, exactly 4 desk items, exactly 3 icons in panel 2, exactly 3 warning icons in panel 3, and exactly 4 workflow steps in panel 4. No photorealistic humans, no extra panels, no watermark, no English replacement for the visible Japanese labels.

Create a square 1:1 painterly editorial portrait in a vintage gouache/oil-pastel illustration style: a young person with {argument name="hair color" default="bright ginger orange"} short cropped hair leans forward in three-quarter view, wearing a {argument name="sweater color" default="coral red-orange"} knit sweater over a soft cream collared shirt. Their hands are clasped prayer-like near the mouth, fingers interlaced and softly modeled, while the entire face is deliberately hidden by a centered opaque matte square censor block in warm peach skin tone, covering from forehead to chin. Surround the figure with lush tropical foliage against a textured dusty blue sky: exactly 3 fully visible lily-like flowers, consisting of 1 large white lily at lower left, 1 large yellow lily at lower right-center, and 1 smaller pink lily on the right; include 1 partial orange-yellow flower cropped at the far right edge. Add broad dark green and blue-green leaves in the foreground, a tall palm frond rising behind the head on the right, and large banana-like leaves framing both sides. Use rough visible brush strokes, grainy paper texture, saturated complementary colors, flattened shapes, warm natural light, no photorealism, no text, no watermark, and a calm contemplative mood. The portrait subject should feel anonymous and modern, with the censor square as a strong graphic element.

Goal: Create a single storyboard sheet for a short cooking video about {argument name="dish" default="Japanese sukiyaki"}, drawn in a nostalgic Showa-era Japanese anime style with warm cel-shaded colors, hand-painted food texture, cinematic close-ups, steam, glossy sauce, and appetizing motion.

Canvas: Extra-wide 21:9 aspect ratio, clean white background, one image containing exactly 12 storyboard panels arranged in a 4-column by 3-row grid. Separate every panel with thin light-gray borders and narrow white gutters. No captions, no panel numbers, no watermark.

Visual style: 1970s–1980s Japanese TV anime food animation, detailed but painterly linework, warm amber kitchen lighting, dramatic close-up compositions, visible steam and bubbling liquid, rich browns and reds, realistic marbled beef texture, soft film-grain feel.

Panel layout and contents, exactly 12 panels:
1. Extreme close-up of thinly sliced marbled beef, vivid red meat with white fat veins filling the entire frame.
2. Close-up of a knife slicing a long white leek on a wooden cutting board, several oval leek slices visible, hand holding the leek.
3. Close-up of a cube of butter melting and foaming in a black iron sukiyaki pan.
4. Thin slices of marbled beef spread across the hot black pan, beginning to sear and sizzle.
5. A small sauce pot pouring dark sukiyaki sauce over beef in the pan, sauce bubbling around the meat.
6. Beef and sliced onions simmering vigorously in glossy brown sauce, heavy steam rising.
7. Tofu cubes and green chrysanthemum leaves added into the pan beside simmering beef and onions.
8. Hands cracking a raw egg into a beige ceramic bowl, bright orange yolk centered in clear egg white.
9. Chopsticks dipping cooked beef into the raw egg mixture in the bowl, golden egg coating the meat.
10. Chopsticks lifting a dripping piece of beef coated in glossy egg against a warm orange background.
11. Side close-up of a person with short dark hair opening their mouth to eat the beef from chopsticks; only the lower face and mouth are visible, steam drifting.
12. Final overhead three-quarter shot of the completed sukiyaki hot pot: beef, leeks, tofu, greens, onions, and noodles arranged in a black pot, steaming generously.

Constraints: Keep all 12 panels clearly separated and in the correct reading order from left to right, top to bottom. Maintain a consistent anime look across panels, use food close-ups rather than wide kitchen shots, and avoid adding any text inside the image.

Create a realistic vertical portrait photograph of a joyful young South Asian woman at golden hour beside a calm river. She stands on a riverside promenade with a metal railing, her body turned partly away from the camera and her head turned back over her shoulder, smiling warmly with closed eyes. She has long wavy dark hair cascading down her back, wears ornate gold jhumka earrings, bangles, and a deep navy blue traditional outfit with sheer draped fabric, delicate small gold embroidery, and wide shimmering gold borders. One hand rests gently on the polished metal railing in the foreground. The river fills the left and center background with a bright golden sun reflection on the water; the right side shows the walkway receding with railings and posts. Background trees and distant riverbank are softly blurred. Use warm sunset backlighting, glowing highlights, shallow depth of field, candid lifestyle photography, natural pose, elegant festive mood, realistic skin and fabric texture, 85mm lens look, soft bokeh, high dynamic range. Keep the composition waist-up to three-quarter length, subject slightly right of center, no text, no watermark.

Create a square editorial fashion portrait of {argument name="character name" default="a young woman"} seated elegantly on a dark polished studio floor, full body visible, posed cross-legged with one knee raised and one leg folded underneath. Keep the face area softly anonymized or unchanged-looking, with no exaggerated expression. She wears a fitted modern black ribbed long-sleeve top with oval shoulder cutouts, loose soft gray washed denim jeans, and chunky gray-and-white platform sneakers. Her hair is {argument name="hair color" default="waist-length wavy milky brown"}, styled half-up with a prominent crown braid and long curled waves falling over one shoulder. Add small dangling silver earrings. The background is a large grayscale mural or photographic backdrop showing an enlarged close-up of braided hair, creating a dramatic monochrome hair-texture composition behind her. Use moody studio lighting, muted gray and black tones, realistic fabric texture, soft shadows, slight floor reflection, high-fashion artistic portrait mood, centered composition, 1:1 aspect ratio. Include exactly 3 main outfit pieces: black cutout top, gray loose jeans, chunky gray-white sneakers. Avoid text, logos, props, extra people, bright colors, or changing the natural seated pose.

Goal: Create a dynamic full-body anime/manga illustration of a comedic fantasy-drama scene featuring {argument name="main character name" default="Hana"} struggling under an absurdly huge boulder while others refuse or fail to help.

Canvas: Vertical 2:3 composition on a clean white background, low-angle perspective that makes the characters look tall and heroic, with the top of the giant boulder cropped by the upper edge of the image.

Subjects and count: Show exactly 5 living subjects: 4 human characters and 1 cat. Human 1 is the central athletic young woman with short messy black hair, wearing a cropped black top, blue bomber jacket, dark high-waisted shorts with chain accessories, thigh strap, black socks, and oversized blue-and-black sneakers; she stands with legs apart and both arms fully raised, palms pressed against a massive cracked stone boulder above her head. Human 2 is a tall blond young man standing to her right, arms crossed, relaxed and unimpressed, wearing a sage-green jacket over a mustard yellow shirt, brown cargo pants with a chain, and rugged green-brown sneakers. Human 3 is a small purple-haired girl at the far lower left, wearing a blue denim outfit and pointing toward the central woman while laughing or accusing. Human 4 is a dark-haired young man at the far lower right, wearing a blue jacket and dark pants, also pointing toward the central woman. Subject 5 is a Siamese-like cat centered near the bottom between the two main adults, cream fur with dark ears, muzzle, paws, and curled tail, bright blue eyes, mouth open in a dramatic meow.

Composition: The central woman and the blond man dominate the frame from shoes to shoulders; their faces are covered by soft rectangular blur blocks as if intentionally censored. The two smaller side characters also have blurred face rectangles. The boulder is enormous, dark gray-brown, rough, cracked, and heavy-looking, with small falling pebbles around it. The central woman is visibly straining, while the blond man remains smugly idle.

Text content: Add exactly 2 speech/handwritten text elements. On the left near the straining woman, handwritten text says: “Help Zack, we’re gonna die!!” with a small pointer line aimed toward her. On the upper right near the blond man, a white speech bubble says: “I’m not helping her”.

Visual style: High-detail contemporary anime illustration with clean ink lines, painterly shading, fashionable streetwear, expressive body language, exaggerated low-angle anatomy, crisp white negative space, subtle ground shadows, and a humorous dramatic tone. Use {argument name="jacket color" default="blue"} for the central woman’s jacket and {argument name="boulder size" default="enormous"} for the stone overhead.

Constraints: Keep exactly 4 humans, exactly 1 cat, exactly 2 text elements, and exactly 1 giant boulder. Do not add a background environment, extra props, extra speech bubbles, logos, or watermarks.

Create a realistic full-body fashion sports portrait of a {argument name="character description" default="fair beautiful young Pakistani woman tennis player"} standing confidently on a professional outdoor tennis court in bright natural daylight. She has long wavy chestnut-brown hair blowing gently in the breeze, a softly anonymized smooth blurred face with no visible facial features, and an elegant relaxed pose with one leg crossed forward. She wears exactly 4 main outfit/accessory items: a {argument name="top color and style" default="black sleeveless athletic crop sports top"}, a {argument name="skirt style" default="premium white high-waisted tennis skirt"}, clean white lace-up sneakers, and small silver jewelry including dangling earrings and a bracelet. She holds exactly 1 tennis racket in her lowered left hand, angled downward with the racket head near the ground. The setting is a blue-and-green hard court with crisp white court lines, a blurred tennis net, green windscreen fencing, and distant stadium seating in the background. Use shallow depth of field, natural sun highlights, soft shadows, realistic skin tones, premium editorial sportswear photography, vertical composition, full-body framing, no text, no watermark.

Goal: Create a cinematic five-panel character photo collage of {argument name="character name" default="a rugged male desert wanderer"} in a vast sand-dune landscape at sunset, with a travel-fashion editorial and film-still mood.

Canvas: Wide horizontal 16:9 collage, approximately 1200×675. Use thin clean white gutters between panels. Warm golden-hour lighting, dramatic orange sunset sky, distant mountains, rippled desert sand, high dynamic range, realistic photography, shallow depth of field where appropriate.

Layout: Use exactly 5 image panels. The left half is one large vertical portrait panel occupying the full height. The right half is divided into four smaller panels in a 2×2 grid.

Panel 1, large left portrait: Close-up seated three-quarter portrait of the man in the desert, cropped from thighs to head. He has dark curly hair, light stubble, tanned skin, and a rugged expression, but his face is covered by a flat opaque rectangular censor block in warm dark brown. He wears a weathered olive-green hooded parka, dusty beige T-shirt, loose cargo pants, wristwatch, bracelets, rings, and a pendant necklace. His hands rest casually on his knees, sunset rim light outlining his jacket and hair.

Panel 2, top middle: Full-body shot of the same man walking toward camera across rippled dunes at sunset, coat tails blowing in the wind, boots kicking up small puffs of sand. His face is also obscured by the same brown rectangular block.

Panel 3, top right: Full-body standing pose on a dune ridge, one hand in pocket, turned slightly to the side, long parka moving in the breeze, warm sky behind him. His face is obscured by a matching brown rectangle.

Panel 4, bottom middle: Seated desert pose, leaning back with one hand on the sand and one knee raised, relaxed and cinematic, dunes and mountains behind him. His face is blurred or covered with a small brown face-obscuring block.

Panel 5, bottom right: Rear view of the man walking away along a dune crest toward the sunset, parka trailing behind, no face visible, silhouetted by orange sky and distant mountains.

Subject details: The character wears {argument name="outerwear" default="a distressed olive-green hooded parka"}, a dusty beige shirt, olive cargo pants, tan lace-up boots, layered bracelets, rings, a dark wristwatch, and a pendant necklace. Hair is {argument name="hair style" default="short dark curly hair"}. Overall palette is {argument name="color palette" default="warm amber sunset, olive green, beige sand, and smoky blue shadows"}.

Visual style: Ultra-realistic cinematic fashion photography, 35mm editorial look, detailed fabric texture, windblown clothing, natural pose variety, crisp sand texture, dramatic sunset backlight, soft bokeh in close-up, consistent character and outfit across all panels.

Constraints: Use exactly five panels and no text. Keep the same character, outfit, desert setting, and sunset lighting consistent throughout. Include the opaque brown rectangular face-obscuring block on all visible frontal faces. No logos, no watermark, no extra people, no vehicles, no modern city elements.

Goal: Create a clean, warm beige Chinese character design poster for {argument name="character name" default="林浅夏"}, a gentle lifestyle blogger, combining a large fashion portrait, expression sheet, pose sheet, item icons, color palette, and biography.

Canvas: Vertical portrait poster, approximately 4:5 ratio, off-white paper background with thin pale beige border, delicate gold-brown line dividers, tiny sparkle and heart doodles, soft editorial magazine layout. Use a warm cream, tan, peach, and coffee color scheme with elegant handwritten accents.

Layout: Split the poster into a large left portrait column and a narrower right information column, with an action-pose section across the lower left and biography/palette sections stacked on the lower right. Keep generous white space and fine hairline separators.

Top left title area: Add the large Chinese headline 「角色设计图」 with a small cursive English subtitle “Character Design”. Beside the portrait, add vertical poetic text 「清风拂面，明朗如光。」 plus small hearts and star doodles.

Main portrait: Show one large waist-up portrait of a young East Asian woman, age {argument name="age" default="25"}, with {argument name="hair style" default="long slightly messy dark brown-black wavy hair"}, warm natural skin, minimal makeup, small gold earrings, a delicate gold pendant necklace, and a cream ribbed sleeveless tank top. She faces forward in soft natural light, elegant and calm, with realistic fashion-photo rendering and subtle watercolor softness.

Right top profile block: Place the large name 「林浅夏」 with a small heart. Under it write 「温柔明朗的生活收藏家」. Include exactly 5 profile lines: 「年龄：25岁」, 「身高：165cm」, 「生日：7月18日」, 「星座：巨蟹座」, 「职业：自由内容创作者」. Add a sticky-note quote card reading: 「生活很简单，开心很重要，请尽情拥抱每一天。——浅夏♡」.

Expression section: Add the heading 「表情设定」 and exactly 6 small bust portraits of the same woman arranged in two rows, each with a rounded beige label. The 6 labels must be: 「开心」, 「生气」, 「困倦」, 「惊讶」, 「害羞」, 「思考中」. Make the hair, outfit, and identity consistent while changing only the facial expression and slight head pose.

Action pose section: Add the heading 「动作设定」 and exactly 3 full-body poses of the same woman in a cream crop tank, beige shorts, white sneakers, and loose natural hair. The 3 labeled poses are: 「跑步」 showing her jogging energetically, 「玩耍」 showing her skipping or playing with bubbles, and 「休闲站姿」 showing a relaxed standing pose holding sunglasses. Use light motion, airy summer styling, and small bubble details around the playful pose.

Object icon section: Add the heading 「物件图标」 and exactly 5 small illustrated icons in one row with labels: 「墨镜」 black sunglasses, 「金色耳环」 gold hoop earrings, 「项链」 gold pendant necklace, 「小白鞋」 white sneakers, and 「小星星」 a small star and heart charm.

Color palette section: Add the heading 「配色方案」 and exactly 6 circular swatches with labels: 「奶油白」, 「米杏色」, 「暖米色」, 「深棕色」, 「蜜桃肤」, 「柔金色」. The palette should visually match cream white, beige, warm tan, dark brown, peach skin, and soft gold.

Biography section: Add the heading 「角色简介」 with a small heart and botanical line art. Include concise Chinese profile text with four labeled lines: 「性格：温柔亲和，开朗乐观，细腻敏感，喜欢发现生活中的小美好。」, 「兴趣：旅行、拍照、写作、听音乐、做手账、收集小物件。」, 「特点：笑容温暖治愈，气质清新自然，拥有轻松感染他人的能量。」, 「简介：林浅夏热爱自由与表达，喜欢用镜头和文字记录生活的点滴。她相信简单的日常里藏着最动人的故事。认真生活，尽情热爱，是她一直坚持的态度。」

Footer: Add a thin decorative line and the centered sentence 「认真生活，尽情热爱，让每一天都闪闪发光。」 with small sparkles at both ends.

Visual style: Soft realistic character-sheet poster, elegant Chinese lifestyle magazine aesthetic, warm daylight, delicate beige typography, subtle watercolor/photoreal blend, clean composition, no harsh shadows, no extra characters, no extra labels, no watermark.

Create a realistic Xiaohongshu and vlog-style outdoor fashion lifestyle photo of a beautiful young East Asian woman posing by a lakeside boardwalk. She stands full-body in the center, leaning lightly against a dark wooden horizontal railing with one hand resting on the rail, one leg crossed elegantly in front of the other, emphasizing a tall slim silhouette, narrow waist, long balanced legs, and graceful posture. She has {argument name="hair color" default="black"} long hair tied into a low ponytail, an oval face with fair skin, delicate facial features, natural light makeup, soft brown eyes, gently arched brows, a small refined nose, and matte coral-pink lips; her expression is gentle and calm, with her head slightly tilted. Dress her in {argument name="top" default="a fitted white zip-up jacket"}, {argument name="bottom" default="black ultra-short shorts"}, and {argument name="shoes" default="transparent strap high-heeled sandals"}. The setting is a quiet lake with soft rippling water, distant layered green mountains, faint buildings across the shore, gray stone pavement, and a dark wooden lakeside fence with vertical posts. Include willow branches hanging naturally from the upper right of the frame. Use an overcast day with diffused natural light, cool muted color grading, shallow depth of field, crisp focus on the subject, softly blurred background, cinematic 8K photorealistic texture, subtle film-like atmosphere, vertical social-media composition, elegant outdoor aesthetic, no text, no watermark.

{argument name="number of people" default="Two people"} eating face to face at {argument name="restaurant" default="Haidilao"}, shot by a girl across the table using an {argument name="device" default="iPhone 17 Pro Max"}

Hyper-realistic close-up portrait of a {argument name="subject" default="woman"} with slightly disheveled, wind-tousled hair, dressed in a {argument name="outfit" default="distressed leather jacket"}. A sweeping white headlight beam cuts across her face from one side, capturing a fleeting moment of brightness against deep shadow. {argument name="setting" default="Urban nighttime"} backdrop, dramatic cinematic motion feel.