GPT Image 2 prompts

Browse GPT Image 2 prompts for fashion portraits and graphic compositions. Compare how each prompt describes its subject and the way the image is arranged.

探索提示词

10,026 条提示词

Social Media Post - Minimalist Luxury Birthday Poster — output

Social Media Post - Minimalist Luxury Birthday Poster

Mr FrancklinGPT Image 2
Create a minimalist and luxurious poster for a milestone birthday, featuring a giant 3D number '{argument name="age" default="[AGE]"}' carved into a high-end ivory white wall. An elegant person is sitting, leaning, or naturally hanging from the number, striking a confident pose. Elegant floral arrangements of white roses, baby's breath, hydrangeas, and pastel blue balloons frame the composition. Soft natural sunlight, high-end editorial photography, modern celebration aesthetic, clean luxury design, subtle shadows, sophisticated typography. At the top, display the name '{argument name="name" default="[NAME]"}' in an elegant serif font, followed by 'CHAPTER {argument name="chapter" default="[AGE]"}' and the quote '365 More Days of Miracles.' Neutral beige and white color palette, high-end event style, dreamy atmosphere, photorealistic, magazine-quality portrait, ultra-detailed textures, luxury birthday photo shoot, Instagram-worthy, shallow depth of field, soft lighting, 8K masterpiece, professional fashion photography.
Comic / Storyboard - Tribal Voodoo Curse Scene — output

Comic / Storyboard - Tribal Voodoo Curse Scene

Michael H. LesterGPT Image 2
black magic
the {argument name="subject" default="tribal chieftain"} moans
in agony
the {argument name="character" default="priestess"} has afflicted me
with a powerful {argument name="curse type" default="voodoo curse"}
E-commerce Main Image - Premium Food Commercial Photoshoot — output

E-commerce Main Image - Premium Food Commercial Photoshoot

AnissaGPT Image 2
Use the uploaded {argument name="subject" default="food/beverage"} image as the main reference. Preserve the exact product identity, appearance, texture, colors, and serving style. Transform it into a premium {argument name="style" default="soft pastel"} studio photoshoot with a bright, clean, modern commercial food photography style. Use a complementary {argument name="background" default="pastel"} background, minimal elegant props, refined plating, soft diffused lighting, realistic shadows, and appetizing natural details. Keep the product as the hero with a balanced composition, negative space, and subtle luxury branding area. Add minimal elegant typography for product name, brand name, and optional CTA. Final output: ultra-realistic 4:5 vertical image, professional food advertisement quality, fresh, polished, premium, and social-media ready.
E-commerce Main Image - Y2K Casual Summer Streetwear — output

E-commerce Main Image - Y2K Casual Summer Streetwear

MoodLock|AI画像の惜しいを直すGPT Image 2
Outfit: a bright summer Y2K casual street outfit with an oversized {argument name="shirt color" default="vivid yellow"} cotton jersey short-sleeve T-shirt featuring a green generalized circular text-style graphic print, styled with a front twisted knot or gathered cropped hem, paired with black mini shorts in lightweight woven fabric with an all-over small white generalized graphic print, white and black chunky sneakers with black laces, spotted or paneled black details, layered synthetic leather or canvas upper, and thick rubber soles; premium fabric texture, realistic garment construction, clean stitching, natural seam tension, believable fabric weight, subtle wrinkles, realistic drape, refined edge finishing, visible seams, real folds, well-made clothing details, high-quality material feel, detailed garment realism, not costume-like, not cheap-looking; add a random mix of playful Y2K {argument name="accessory material" default="silver-tone"} accessories such as bangles, rings, and small earrings that naturally match the outfit without fixing exact pieces.
E-commerce Main Image - Y2K Casual Street Hoodie Outfit — output

E-commerce Main Image - Y2K Casual Street Hoodie Outfit

MoodLock|AI画像の惜しいを直すGPT Image 2
Outfit: cute Y2K casual street outfit with an oversized {argument name="hoodie color" default="off-white"} cotton fleece hoodie worn in a mini-length silhouette, front pouch pocket, relaxed long sleeves, ribbed cuffs and hem, and a generalized chest graphic print, layered with a {argument name="vest color" default="light-blue"} washed denim-like quilted puffer vest featuring an all-over generalized monogram-style pattern, high padded collar, open front, believable padding volume, subtle yellow lining accent, visible seams, real folds, natural seam tension, realistic drape, clean stitching, premium fabric texture, and refined edge finishing; a very short beige to light-brown mini bottom subtly visible beneath the hoodie hem; white chunky platform lace-up sneaker boots with faux-fur ankle trim, thick rubber sole, metal eyelets, small generalized patch-like side details, high-quality material feel, detailed garment realism, not costume-like, not cheap-looking; add a random mix of delicate casual Y2K {argument name="accessory tone" default="silver-tone"} and white-tone accessories that match the outfit without fixing exact pieces.
E-commerce Main Image - Y2K Street Ribbed Tank Top Outfit — output

E-commerce Main Image - Y2K Street Ribbed Tank Top Outfit

MoodLock|AI画像の惜しいを直すGPT Image 2
Outfit: a casual Y2K street outfit with a fitted {argument name="top color" default="gray"} ribbed knit tank top or camisole-style top featuring a scoop neckline, compact silhouette, premium fabric texture, realistic garment construction, clean stitching, natural seam tension, believable fabric weight, subtle wrinkles, realistic drape, and refined edge finishing, paired with black to charcoal vertical pinstripe balloon mini shorts / bloomer-style mini skort in lightweight woven fabric with rounded gathered volume, visible seams, real folds, and detailed garment realism, white ribbed knee-high socks, and a {argument name="hat color" default="burgundy"} cotton twill baseball cap with metal hardware, no bag; high-quality material feel, well-made clothing details, not costume-like, not cheap-looking; add a random mix of delicate casual Y2K silver-tone accessories such as rings, bracelets, and layered necklaces that naturally match the outfit without fixing exact pieces.
Product Marketing - Red Circle Luxury Skincare Fashion Ad — output

Product Marketing - Red Circle Luxury Skincare Fashion Ad

Cherry 2.OGPT Image 2
Create a high-fashion luxury skincare commercial advertisement on a square studio canvas. A slim young Asian female model with long flowing black hair walks confidently out of a giant minimalist red circular frame positioned behind her, centered like a bold graphic halo. She is frozen in a dynamic runway stride with one leg extended forward and the other behind, creating a strong elegant silhouette. She wears a vibrant flowing red one-shoulder chiffon dress with a high slit, ruched bodice, translucent fabric layers, and wind-swept skirt billowing dramatically to the left, paired with red ankle-strap stiletto sandals. Her right arm reaches toward the viewer holding a red cylindrical luxury skincare bottle labeled {argument name="product label" default="SK-II"}, making the bottle prominent in the foreground; her left arm trails naturally behind. Use a clean light gray studio background, glossy floor contact shadows, crisp commercial lighting, subtle shadows from the circular frame, premium beauty-ad styling, red-on-gray minimalist palette, photorealistic editorial fashion photography, sharp focus, elegant confidence, no extra text, no logo except the bottle label.
Profile / Avatar - Cinematic Five-Expression Portrait — output

Profile / Avatar - Cinematic Five-Expression Portrait

Cherry 2.OGPT Image 2
Create a cinematic black-and-white studio portrait collage of the same {argument name="character description" default="handsome young adult man with dark swept-back hair, light stubble, strong jawline, and expressive eyes"} shown in exactly five narrow vertical panels separated by thin glowing white divider lines. Use a horizontal 4:3 canvas with a deep black background and dramatic low-key lighting, high contrast, crisp skin texture, glossy hair highlights, and a monochrome editorial fashion look. The man wears a plain black crew-neck shirt in every panel, framed from upper chest to top of head, with each face large and close to the camera. Show exactly five expressions from left to right: 1) intense serious stare with furrowed brow looking directly forward, 2) warm closed-mouth smile looking forward, 3) broad open laugh with eyes slightly squinting, 4) thoughtful skeptical look glancing upward to the side with one hand resting on his chin, 5) calm neutral direct gaze with relaxed face. Keep the same identity, hairstyle, clothing, lighting style, and scale across all panels. Make the image feel cinematic in every pixel, like a dramatic actor expression study or emotion contact sheet. No text, no logos, no color, no extra people, no props other than the hand in the fourth panel.
Social Media Post - Eight Mini Me Family Portrait — output

Social Media Post - Eight Mini Me Family Portrait

Abkr SadiqGPT Image 2
Create an ultra-detailed studio “Mini Me family portrait” featuring one real-looking central person and exactly eight miniature childlike clones of him, all sharing the same facial identity, skin tone, short black haircut, expressive dark eyes, eyebrows, nose, lips, ears, smile shape, and overall facial proportions. The central subject is a young Black man named {argument name="character name" default="the subject"}, seated front-and-center in a wicker peacock chair, facing the camera with a calm neutral expression. He wears a navy T-shirt with a distressed Texas-flag-style design, ripped blue jeans, white high-top sneakers, and a wristwatch. Around him are exactly 8 Mini Me versions: 1) a smiling mini sitting on the top left edge of the chair back, legs dangling; 2) a mini standing at far left near a potted palm with arms crossed and a confident side glance; 3) a mini sitting on stacked vintage brown suitcases at bottom left, looking upward thoughtfully; 4) a mini leaning against the man’s left thigh with chin resting on hands, gazing up admiringly; 5) a mini sitting on a small wooden stool at right, smiling toward the man; 6) a mini standing behind the stool on the right, holding a camera up as if photographing the scene; 7) a mini lying on his stomach on the floor at bottom right, drawing a pencil portrait in a spiral sketchbook; 8) a mini partially perched behind the central subject’s left shoulder/upper chair area, playful and cheerful. Dress every Mini Me in matching navy Texas-flag T-shirts, blue jeans, and white sneakers, scaled proportionally, with cute toddler-like bodies but identical recognizable face and hairstyle. Use a warm professional portrait studio setting with a mottled brown textured backdrop, woven jute rug, wicker chair, potted palm on the left, wooden side table on the right with a small plant, vintage leather suitcases, a pencil cup with colored pencils, and the open sketchbook showing a face drawing. Lighting should be cinematic warm softbox lighting, gentle shadows, high realism, sharp facial detail, rich skin tones, subtle depth of field, cozy earthy color palette, vertical 4:5 composition, no text, no watermark, no extra characters, and make all nine figures feel physically present in the same scene.
Profile / Avatar - Sepia Vintage Intellectual Portrait — output

Profile / Avatar - Sepia Vintage Intellectual Portrait

Chain LoaderGPT Image 2
Create a sepia-toned vintage photographic portrait of {argument name="subject" default="a middle-aged intellectual man"} leaning forward on a rough wooden railing, shown from the chest up in a three-quarter view facing left. He has a weathered face, deep forehead lines, a serious contemplative expression, messy windblown light hair, and small round wire-rim glasses. He wears {argument name="clothing" default="a loose white linen shirt with rolled sleeves and an open collar"}, slightly wrinkled and sunlit. His forearms rest across the rustic railing in the foreground. Use dramatic natural light from the upper left, with the face partly illuminated and the background falling into deep black shadow, suggesting an outdoor porch or rural setting. Render as an early 20th-century sepia gelatin silver or albumen print: warm brown monochrome tones, high contrast, soft focus at the edges, visible film grain, dust specks, scratches, vignetting, uneven exposure, aged paper texture, and slightly faded archival wear. Composition should be intimate and documentary, with the subject placed slightly right of center, large dark negative space behind him, no modern objects, no text, no watermark.
Social Media Post - Golden Hour Riverside Portrait — output

Social Media Post - Golden Hour Riverside Portrait

Cherry 2.OGPT Image 2
Create a cinematic lifestyle portrait of {argument name="character name" default="a young East Asian woman"} leaning peacefully on a riverside bridge railing during golden hour. She has a short black bob haircut with soft wind-blown strands, delicate facial features, fair skin, and round clear eyeglasses. She is looking slightly upward with her eyes gently closed and a subtle relaxed smile, conveying quiet peace and solitude. She wears an oversized {argument name="sweater color" default="light gray"} chunky cable-knit sweater with long sleeves covering part of her hands, plus a small delicate necklace. Her arms rest casually on a weathered wooden bridge rail, with metal vertical railings receding into the distance on the right. The background shows a calm river reflecting warm sunset light, a softly blurred urban skyline with tall buildings, and a distant hill with a slender tower, all in creamy bokeh. Use warm backlighting, soft rim light on her hair and sweater, shallow depth of field, natural skin texture, realistic photography, 85mm lens look, cinematic color grading, serene atmosphere, no text, no watermark.
Infographic / Edu Visual - Minimalist Tokyo Morning Poster — output

Infographic / Edu Visual - Minimalist Tokyo Morning Poster

小小东GPT Image 2
Create a cinematic minimalist travel poster for {argument name="city name" default="Tokyo"}, using a vast dark navy-blue architectural facade as the background and a tiny solitary pedestrian at the bottom for scale. The composition should feel grand, cold, clean, quiet, and ritualistic, with a monumental wall, sharp vertical seams, a narrow slit of reflected city glass near the center, and pale concrete pavement along the bottom. Use a deep blue monochrome color grade, high contrast sunlight and shadow, elegant editorial poster aesthetics, fine white typography, generous negative space, and a calm early-morning atmosphere.

Layout: horizontal 16:9 poster. Left third contains small structured guide text and numbered items over the building wall; right third contains the main destination title. Top left reads “G O O D   M O R N I N G” in widely spaced small serif capitals. On the left, show exactly 3 numbered information blocks: “01” in a rounded rectangle followed by “地标|东京塔与山手线”; “02” on a thin horizontal divider followed by “气息|站台广播、夏祭薄风、清晨街角”; “03” inside a thin double-line circle followed by “色彩|藏蓝 / 提灯红 / 和纸白”. Bottom left shows exactly 2 small icon badges: a diamond icon labeled “收藏价值 / COLLECTIBLE VALUE” and a classical column icon labeled “美学价值 / AESTHETIC VALUE”.

Right side typography: place huge vertical white Japanese characters “东京” as the dominant title, with “T O K Y O” set vertically beside it in spaced serif letters. Under the title, add a thin horizontal line with a small red seal emblem at center, then small text “东京” and “とうきょう”. Below that add the date “{argument name="date" default="2026-07-10"}” and the sentence “早安,愿你在秩序与霓光之间,保持清醒与温柔。” On the far right edge, include one ornate vertical label frame marked “04” with the text “漫游|先去便利店买一瓶冰水”. Near the lower right, include one rectangular outlined box marked “05 记忆|收藏日常里的 精准与留白”.

Subject details: the only person is a tiny figure in dark clothing walking across the pale plaza near the lower center-left, casting a long subtle shadow. No cars, no crowds, no clutter.

Style constraints: ultra-clean PPT-like grid logic, refined Japanese editorial design, photorealistic architectural background, crisp white overlays, one small red accent seal, no watermark, no extra text beyond the specified poster elements. Use {argument name="main color palette" default="deep navy blue, paper white, small vermilion red accent"} and keep the mood {argument name="mood" default="quiet, majestic, cool, clean, ceremonial"}.
Social Media Post - Sunlit Orchard Girl With Apples — output

Social Media Post - Sunlit Orchard Girl With Apples

道不生AI笔记GPT Image 2
Create a vertical 2:3 photorealistic portrait of a young East Asian woman standing in a lush green orchard or garden, framed from the waist up and centered. She has {argument name="hair color" default="dark brown-black"} hair with soft wispy bangs and one thick side braid falling over her right shoulder, decorated with a small silver star-shaped hair clip on the viewer’s right side of her head. She faces the camera directly with a calm, slightly wistful neutral expression, clear fair skin, soft natural makeup, and large gentle eyes. She wears a sleeveless pastel plaid summer dress in muted cream, green, blue, and pink tones, with a simple gathered waist. Both hands are held at chest height, gently cradling exactly 2 pale green apples side by side, each apple with a small visible stem. The background is dense sunlit foliage with large green leaves, shallow depth of field, dreamy circular bokeh, and warm backlighting from the upper left; include strong golden rim light glowing around her hair and shoulders, soft lens flare, and a nostalgic film-photography look. Use natural outdoor lighting, 35mm or 50mm portrait lens feel, delicate skin texture, soft highlights, subtle grain, high dynamic range, and a serene summer atmosphere. Keep the composition clean with no extra people, no extra fruit, no text overlays, no logos, and no watermarks.
Social Media Post - Gritty Industrial Fashion Portrait — output

Social Media Post - Gritty Industrial Fashion Portrait

周防義経GPT Image 2
Create a photorealistic vertical fashion portrait of an adult Japanese woman, {argument name="model age" default="early 20s"}, standing in a dark industrial studio against a stained raw concrete wall with visible seams, bolts, grime, and moody shadows. She has a short chin-length bob haircut with wispy bangs, {argument name="hair color" default="dark brown"}, clear natural makeup, large expressive eyes, and a gentle slight smile while looking directly at the camera. Pose her with a relaxed tilted torso and one shoulder dropped, giving a confident casual gravure-style fashion pose. Outfit: an off-shoulder sheer black mesh long-sleeve cropped top layered over a black bandeau, exposing the midriff, paired with oversized dark olive cargo pants, a black double-buckle grommet belt, and a dangling silver chain at the hip. Lighting should be dramatic and directional from the upper left, creating strong highlights on the face and abdomen with deep shadows around the edges. Use realistic skin texture, natural proportions, shallow depth of field, high-detail fabric transparency, cinematic contrast, and a gritty urban editorial mood. No text, no logos, no watermark.
Product Marketing - Symmetric Roland Garros Paris Poster — output

Product Marketing - Symmetric Roland Garros Paris Poster

小小东GPT Image 2
Create a vertical cinematic sports event poster for {argument name="tournament name" default="Roland Garros"} in Paris, using a perfectly centered axial symmetry composition. A red clay tennis court fills the lower half, with crisp white court lines and a tiny lone tennis player in white standing exactly on the center line facing away toward the horizon. On both sides are packed stadium stands with small crowds, dark sponsor boards, and French flags, mirrored left and right. In the distance, place the Eiffel Tower centered behind the court, rising into a pale blue morning sky with soft clouds and a bright airy atmosphere. Add a surreal symmetrical procession of yellow tennis balls floating in the sky along the central perspective axis: exactly 12 balls total, with 2 very large balls near the upper left and upper right foreground labeled “ROLAND GARROS PARIS,” 2 medium balls below them, and 8 progressively smaller balls receding toward the Eiffel Tower, evenly balanced on both sides. Use refined luxury travel-poster aesthetics, realistic photo-composite lighting, elegant minimal typography, warm clay orange contrasted with sky blue and tennis-ball yellow. Top text layout: small spaced “2025” centered at the very top; huge wide-letterspaced “ROLAND GARROS”; smaller centered “PARIS” with thin horizontal lines on both sides; date line “MAY 19 - JUNE 8”; orange uppercase tagline “THE CLAY. THE LEGEND.”; larger orange uppercase headline “THE MOMENT.”; small white uppercase line “TICKETS ON SALE NOW.” Bottom center: a white Roland Garros circular logo on the clay, and beneath it the slogan {argument name="bottom slogan" default="ONE COURT. COUNTLESS STORIES."}. Keep all typography clean, uppercase, widely tracked, premium, and centered. No extra people besides the single player, no extra logos except the tennis-ball labels and bottom circular logo, no watermark.
Comic / Storyboard - Sunset Programmer Bedroom Studio — output

Comic / Storyboard - Sunset Programmer Bedroom Studio

老范讲故事🎙️GPT Image 2
Create a highly detailed watercolor-and-ink illustration of a messy programmer’s bedroom studio at sunset, viewed from inside the room toward a huge floor-to-ceiling window. The scene shows a warm orange city skyline and highway traffic outside, with the sun low between high-rise buildings and long glowing reflections. Inside, make the room cramped, lived-in, and cluttered: an unmade bed on the left with rumpled blue-gray striped bedding, a small bedside table with books and a lamp, framed wall art, shelves with books and storage, hanging plants, and scattered newspapers on the wooden floor. In the center foreground place a black mesh office chair facing a large wooden desk. On the right side of the desk include exactly 2 computer monitors showing dense programming code, with the larger monitor angled on the far right and the smaller monitor behind it near the window. Add exactly 2 large black studio speakers, one near the monitors and one prominent at the far right, plus a warm articulated desk lamp shining on the workspace. Cover the desk with developer clutter: a mechanical keyboard, cables, pens in a cup, notebooks, takeaway food containers, crumpled paper, a coffee mug, small electronics, tools, and headphones. Fill the upper right wall with crowded bookshelves. Use cinematic golden-hour lighting, deep shadows, amber highlights, urban melancholy, expressive brush texture, rough paper edges, visible watercolor washes, fine ink linework, high detail, no people, no readable brand names, no watermark.
Product Marketing - NESCAFÉ Coffee Map Poster — output

Product Marketing - NESCAFÉ Coffee Map Poster

Sara | 사라 Al ✦ 🎥✍️GPT Image 2
Goal: Create a premium NESCAFÉ coffee campaign poster with a realistic hero coffee cup integrated into an antique hand-drawn coffee-map illustration.

Canvas: Vertical 4:5 poster on warm ivory parchment paper, centered composition, high-end editorial advertising feel, no border.

Main subject: A top-down photorealistic espresso cup in the exact center, with a glossy bright red saucer and matching red handle facing right. The cup has a thick white inner rim and dark coffee with rich golden-brown crema, tiny bubbles, foam speckles, and a soft reflective highlight. The cup sits over a circular coffee-stain ring and pale watercolor wash.

Branding and text: At the top center place the large NESCAFÉ logo, dark coffee-brown with the small red accent over the final letter, using the recognizable brand style. Directly below it write {argument name="tagline text" default="THE ART OF COFFEE"} in spaced uppercase serif letters. Under the tagline add a small horizontal ornamental divider with a single coffee bean at the center. At the bottom center add the small credit text {argument name="signature text" default="@sara4a"} with thin decorative horizontal lines on both sides.

Illustrated map elements: Surround the cup with sepia ink drawings and light coffee-watercolor splashes, arranged as exactly 8 distinct illustrated feature groups: 1) a tall coffee plant branch with leaves and coffee cherries on the upper left, 2) a small transparent glass coffee mug on the left middle, 3) a mountain range with rising sun and three flying birds on the upper right, 4) a small rustic farmhouse on the right middle, 5) leafy coffee branches with berries on the right side, 6) a vintage wooden manual coffee grinder on the lower right, 7) a coffee plantation field with rows, hills, and tiny houses along the bottom, 8) a large coffee blossom and branch cluster on the lower left. Scatter a small number of loose coffee beans around these drawings.

Visual style: Mix crisp photorealistic product photography for the red cup with delicate sepia botanical engraving, vintage travel-map line art, coffee stains, beige watercolor splatters, and warm brown monochrome illustration. Use premium coffee tones: cream, tan, caramel, espresso brown, and vivid glossy red.

Constraints: Keep the composition symmetrical and centered, make the cup the dominant focal point, keep all illustration details behind and around the cup, use clean readable branding, avoid extra slogans or unrelated objects, and maintain a polished luxury advertising look.
Game Asset - Cyberpunk Megacity Hacker Scene — output

Game Asset - Cyberpunk Megacity Hacker Scene

小海豚笔记 (AI & 副业探索)GPT Image 2
Create a cinematic cyberpunk key visual for a social media banner or concept-art poster, showing {argument name="character name" default="a young urban hacker"} standing in the left foreground with their back and side profile visible, holding a glowing smartphone whose screen is clear and readable enough to suggest a futuristic city-control app. The character has messy dark hair, wears a glossy black tech jacket with small neon text patches and luminous blue-magenta accents, and looks out over a vast rainy megacity at night. The city should fill the background with towering ultra-thin skyscrapers, dense windows, elevated bridges, flying vehicles, hovering drones, and huge holographic billboards; include exactly 3 prominent aerial vehicles in the sky, exactly 1 large circular suspended structure in the upper left, and exactly 2 large holographic female-face displays: one giant translucent face projected across the central skyscraper and one smaller face billboard on a right-side tower. Use a dramatic wide-angle composition, low rooftop viewpoint, wet reflective surfaces, deep blue and violet atmosphere, neon cyan and magenta highlights, mist, rain-slick reflections, volumetric light beams, and a moody storm-cloud sky with a faint purple dawn glow on the horizon. Leave the right and upper central skyline visually rich but not covered by captions, with no watermark, no logo, and no extra readable text beyond small futuristic interface details. Style: high-detail sci-fi concept art, photorealistic lighting, sharp architectural detail, immersive scale, 16:9 horizontal canvas.
Infographic / Edu Visual - Hand-Drawn Tomato Egg Recipe Infographic — output

Infographic / Edu Visual - Hand-Drawn Tomato Egg Recipe Infographic

Gorden SunGPT Image 2
Goal: Create a cute hand-drawn Chinese recipe infographic in an open notebook, teaching how to make {argument name="dish name" default="西红柿鸡蛋"}.

Canvas: Wide 16:9 horizontal composition, like a two-page spread of a cream-colored grid-paper notebook with rounded page corners, faint beige graph lines, soft pencil outlines, and a warm watercolor texture.

Layout: At the top left, write the large handwritten Chinese title “如何制作西红柿鸡蛋” with a pale yellow highlighter stroke underneath. Across the upper middle are exactly 6 taped ingredient sticky notes, left to right: 1) red note labeled “西红柿” with a tomato drawing, 2) yellow note labeled “鸡蛋” with two eggs and a paperclip, 3) green note labeled “葱” with green onions, 4) blue note labeled “油” with a small oil bottle, 5) purple note labeled “盐” with a salt shaker, 6) pink note labeled “糖” with a sugar jar. In the upper right, show the finished tomato-and-egg dish on a white plate on a wooden tray, with a small bowl of chopped scallions, a spoon, sparkles, a red heart, and handwritten text “酸甜可口!”.

Recipe steps: Along the lower half, place exactly 5 torn-paper recipe cards connected by blue arrow lines. Card 1 has a red border and label “1. 切”, showing sliced tomatoes on a cutting board, a cleaver, and a small bowl of chopped scallions. Card 2 has a yellow border and label “2. 炒鸡蛋”, showing scrambled eggs in a black frying pan over small orange flames. Card 3 has a green border and label “3. 炒西红柿”, showing tomato slices cooking in a frying pan. Card 4 has a blue border and label “4. 混合”, showing tomatoes and eggs mixed together in the pan. Card 5 has a pink border and label “5. 调味”, showing the finished mixture in the pan, two small spoons above it, and a rounded label with a green check mark reading “完成!”.

Decorative details: Add a small smiling cartoon girl peeking from the left side near the first step card, holding a pointer. Add potted plants on the far left margin, scattered chopped scallions, tiny hearts, sparkles, a wooden spoon on the lower right, and a small bowl with spoon on the lower left. At the bottom center, write the handwritten slogan “美味家常菜,简单又健康!” with red hearts and leafy flourishes.

Visual style: Soft pastel watercolor, colored pencil sketch lines, kawaii cooking illustration, scrapbook collage aesthetic, masking tape on the notes, light shadows, cozy homemade feel, neat legible Chinese handwriting.

Constraints: Use exactly 6 ingredient notes and exactly 5 step cards; keep all Chinese text legible and in the positions described; no photorealism, no extra recipe steps, no watermark, no logo.
E-commerce Main Image - Refreshing Citrus Coconut Cloud Ice Poster — output

E-commerce Main Image - Refreshing Citrus Coconut Cloud Ice Poster

小小东GPT Image 2
Goal: Create a refreshing vertical summer beverage poster for {argument name="drink name" default="青柠椰云冰饮"}, with a transparent icy aesthetic, pale cyan palette, and premium minimalist Chinese typography.

Canvas: Tall 9:16 poster, airy high-key lighting, soft sky-blue and white background, faint clouds, sparkling mist, crushed ice, and a glossy water surface reflection at the bottom. Add a soft palm-leaf shadow in the upper right to suggest tropical summer light.

Main subject: Place one large clear plastic cup slightly right of center in the lower-middle of the poster. The cup contains a translucent light-aqua iced drink with visible bubbles, condensation droplets, crushed ice at the top, and exactly 1 round lime slice garnish leaning on the rim. Make the cup rim and walls transparent, glossy, and realistic, with cool highlights and refractions.

Cup label: Add a rounded white label on the cup with teal printing. The label should contain exactly 1 main Chinese drink name, the English subtitle “CITRUS COCONUT CLOUD ICE,” 1 thin divider line, and exactly 3 small line icons: a citrus slice, a coconut with palm tree, and ice cubes. Add small ingredient text at the bottom reading “青柠 | 椰子 | 冰爽”.

Typography and poster text: On the left side, create a very tall vertical headline in thin teal Chinese characters reading {argument name="vertical headline" default="青柠椰云冰饮"}. Near the top-left, add a small vertical slogan reading {argument name="slogan" default="清爽椰云 冰凉入心"} with exactly 1 small lime-slice icon above it and a short hanging dot-line ornament. Near the lower-left, add exactly 3 stacked English words: “Citrus”, “Coconut”, “Cloud Ice”.

Stamp detail: On the right side above the cup, add exactly 1 circular teal seal stamp with a snowflake icon in the center and small Chinese text around the ring suggesting summer recommendation and ice-cool refreshment.

Visual style: Photorealistic product-ad rendering mixed with clean graphic design, elegant thin-line teal typography, ultra-translucent glassy ice, soft bloom, watery reflections, low contrast, generous negative space, fresh coconut-lime summer mood, no people, no clutter, no extra products, no watermark.
Comic / Storyboard - Neon Sunset Kimono Girl — output

Comic / Storyboard - Neon Sunset Kimono Girl

空想製作所/AI4コマ漫画GPT Image 2
Create a cinematic anime illustration of {argument name="character name" default="a young Japanese woman"} in an open-air traditional corridor at sunset, turning back over her shoulder with a gentle sleepy smile as if saying good night. She wears an elegant dark navy-black kimono with vivid orange lining and stylized cloud patterns, tied with a large obi bow, with long flowing {argument name="hair color" default="black"} hair swept dramatically by the wind and decorated with small blue-purple flowers. The scene is a fusion of traditional Japanese architecture and neon cyberpunk lighting: wooden pillars and railings frame a wide view of a huge orange sun setting behind layered mountains, clouds, and a reflective river or bay, while glowing cyan-blue rectangular lantern signs hang overhead along the corridor. Use a wide 16:9 composition, dynamic diagonal perspective, the character positioned slightly right of center, her kimono sleeves and ribbons spreading across the foreground. Fill the air with exactly three visible types of windblown decorative elements: pink flower petals, red-orange leaf-like fragments, and long purple ribbon streamers. Lighting should be highly dramatic with warm sunset orange from the left and cool electric blue-magenta neon from above and right, creating glossy highlights, deep shadows, rim light on hair strands, and saturated color contrast. Style: ultra-detailed Japanese anime key visual, painterly digital art, sharp linework mixed with atmospheric brush texture, high contrast, luminous glow, romantic nocturne mood, no text, no watermark, no modern signage lettering.
Game Asset - Full-Bleed Crane Park Key Art — output

Game Asset - Full-Bleed Crane Park Key Art

ヤノGPT Image 2
Using the provided reference image, isolate and regenerate only the central park illustration as a full-bleed wide cinematic artwork. Remove all presentation-board elements, typography, swatches, maps, thumbnails, labels, margins, and paper background. Expand the central scene to fill a {argument name="aspect ratio" default="16:9 landscape"} canvas while preserving the same painterly concept-art style, dusk blue-and-gold lighting, crane-park environment, child figure, swing/crane landmark, flowers, trees, and atmospheric urban skyline. Increase clarity and detail without making it look photorealistic; keep the textured gouache/pastel brushwork and hand-painted grain. Make it suitable as a clean environment key art image with no text or UI elements.
E-commerce Main Image - Pastel Peach Lychee Rose Parfait — output

E-commerce Main Image - Pastel Peach Lychee Rose Parfait

菜々花GPT Image 2
Create a soft, elegant food photography image of a {argument name="dessert type" default="peach and lychee rose parfait"} in a tall clear fluted parfait glass, centered on a pale white marble tabletop with subtle gray veining. The parfait has exactly 5 visible layered bands from bottom to top: translucent coral-pink jelly, thick ivory cream, chunky pink peach compote, another thick ivory cream band, and a raspberry-red sauce band just below the rim. On top, include exactly 7 main toppings: one large peach-colored rose-shaped sorbet or mousse swirl with delicate gold flakes, three glossy peeled lychees with small orange centers, one mound of dark pink crumble, and two small pink edible blossoms. Use a refined pastel palette of blush pink, peach, ivory, and coral. Surround the glass with a dreamy shallow-depth-of-field setting: a blurred bouquet of pink flowers in the upper left, one peach-pink rose-like dessert or flower at the left edge, exactly 2 round pink confection balls on the right, and exactly 3 loose petals on the tabletop in the foreground and right side. Lighting should be natural, diffused, and high-key, with gentle highlights on the glass and lychees, a luxurious spring dessert styling, realistic textures, no people, no text, no logo, no utensils, and no harsh shadows. Use {argument name="camera angle" default="slightly elevated three-quarter view"}, {argument name="color palette" default="pastel blush pink and ivory"}, and {argument name="background style" default="softly blurred floral marble tabletop"}.
Comic / Storyboard - Anime Character Generation Workflow — output

Comic / Storyboard - Anime Character Generation Workflow

ジジはAI漫画部族GPT Image 2
Goal: Create a dark UI-style comparison canvas showing three generated anime-image results connected in a workflow graph, as if viewed inside an image generation tool.

Canvas: Wide 16:9 screenshot composition with a nearly black dotted-grid background. Use subtle blue connector lines and small circular image-node icons between panels. The layout contains exactly 3 rounded rectangular panels: 1 large character sheet panel on the lower left, 1 cinematic panel on the upper right, and 1 cinematic panel on the lower right.

Layout: Place the lower-left panel as the source reference node, labeled in small white Japanese text above it: 「画像生成ツール #2」. From its right side, draw a branching blue connector line to the two right panels. Label the upper-right panel 「画像生成ツール #10」 and the lower-right panel 「画像生成ツール #1」. Add tiny yellow circular refresh-style icons near the top-right corner of the right-side panels. Keep all panels with thin gray borders, rounded corners, and faint drop shadows.

Left panel content: Show a detailed anime character design sheet for {argument name="character name" default="Aban"}, a tan-skinned young male adventurer with medium-length brown dreadlocks, athletic build, confident expression, sleeveless dark tunic, brown leather vest with straps, fingerless gloves, dark pants, boots, pendant necklace, and a flowing red scarf. The sheet contains exactly 3 full-body turnaround views in the center: front view, side view, and back view. It also contains exactly 6 expression portrait boxes on the right arranged in a 2-column by 3-row grid: normal smile, broad grin, serious, surprised, angry, and sad. On the left side of the sheet, include a larger bust portrait under a large Japanese name title 「アバン」, with small Japanese profile text blocks and stat-like labels beneath. Use beige parchment paper, clean black manga-style line art, and muted watercolor anime coloring.

Upper-right panel content: Show the same character in a cinematic desert scene at sunset, framed from waist up with a dramatic wide-angle lens. He smiles confidently while reaching one hand toward the camera, fingers foreshortened, with a glowing orange fire-magic spark forming near his palm. His red scarf blows behind him. The background has cracked desert ground, distant tents, low mountains, warm orange clouds, and golden sunlight. Add a white Japanese caption along the lower edge of the image: 「このキャラ 火花魔術を使うカットを生成して、広角レンズ hand to camera な構図」.

Lower-right panel content: Show a more dynamic anime action version of the same scene, closer to the camera and more intense. The character reaches one large foreshortened hand directly toward the viewer, with a bright circular burst of fiery magic at the palm, orange sparks flying across the frame, blue sky behind him, desert cliffs in the distance, and his red scarf whipping dramatically. Use high-energy perspective, strong motion, vivid lighting, and sharp anime rendering. Repeat the same white Japanese caption along the bottom: 「このキャラ 火花魔術を使うカットを生成して、広角レンズ hand to camera な構図」.

Visual style: Sleek AI-tool screenshot aesthetic, dark interface, crisp panel edges, small UI labels, anime fantasy illustration inside each panel. The character should remain consistent across all 3 panels: {argument name="hair color" default="brown dreadlocks"}, tan skin, red scarf, leather adventurer outfit, friendly but heroic personality. Use the magic effect color {argument name="magic color" default="orange fire"} and the environment {argument name="setting" default="desert at sunset"}.

Constraints: Include exactly 3 main image panels, exactly 3 full-body turnaround views, exactly 6 expression portraits, exactly 2 right-side cinematic action images, and exactly 1 branching connector from the left panel to the two right panels. Keep the Japanese text legible enough to read where specified, but do not add extra large titles outside the UI labels.
按模型、风格或主题浏览

寻找下一个方向

ASTORIE CANVAS

让灵感,继续生长。

在同一个创作空间,连接提示词、参考素材和生成结果。

打开创作画布 ↗