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

9,926 条提示词

Infographic / Edu Visual - Technical Product Infographic Overlay — output
Social Media Post - Serene Four-Panel Woman Portrait — output
Comic / Storyboard - Hand-drawn travel journal illustration — output
Product Marketing - Luxury East Asian Fashion Portrait — output
Product Marketing - Editorial Fashion Menswear Portrait — output
Product Marketing - Cinematic Portrait in Misty Kashmir — output

Product Marketing - Cinematic Portrait in Misty Kashmir

HeisenLegacyGPT Image 2
Ultra-realistic cinematic editorial photograph of a strikingly handsome {argument name="subject" default="young man in his early 20s"} standing on a quiet mountain road in {argument name="location" default="Kashmir"} during a misty early morning. He has thick, naturally wavy jet-black hair, strong well-defined eyebrows, deep expressive dark-brown eyes, a straight sculpted nose, prominent cheekbones, a sharp masculine jawline, and subtle natural stubble. He wears a sophisticated traditional-modern Kashmiri outfit — a beautifully tailored charcoal pheran layered over a crisp cream kurta, slim dark trousers, and elegant leather boots, blending timeless Kashmiri heritage with contemporary fashion. He stands confidently with a calm, thoughtful expression, one hand resting naturally in his pocket. Behind him, majestic Himalayan mountains, pine-covered slopes, and distant valleys gradually disappear into thick layers of soft white morning fog. Dew-covered grass, subtle mist drifting across the road, muted earthy tones, cool morning atmosphere, soft diffused sunrise light breaking gently through the fog, cinematic depth, realistic fabric texture, natural skin pores, atmospheric perspective, shallow depth of field, 85mm lens, photorealistic, 8K HDR, luxury fashion editorial aesthetic, dramatic yet peaceful composition, authentic Kashmir landscape, no artificial-looking face, no text, no watermark. Vertical 9:16 composition.
使用此提示词 ↗
Social Media Post - Korean Inspired Summer Lifestyle Portrait — output

Social Media Post - Korean Inspired Summer Lifestyle Portrait

AqsaGPT Image 2
Ultra-realistic outdoor lifestyle portrait of a {argument name="subject" default="young woman standing on a quiet tree-lined street"} on a bright summer day. She has long, silky dark-brown hair with soft natural waves, subtle curtain bangs, and loose strands framing her face. She has delicate youthful facial features, clear natural skin, soft brown eyes, subtle makeup, and a cute playful expression with gently pursed lips.\nShe is wearing a {argument name="clothing" default="fitted white sleeveless square-neck dress with a clean minimalist style"}. A soft beige knitted cardigan is draped casually over one arm. She carries a small elegant white quilted handbag with a gold chain and gold hardware. She wears a delicate pendant necklace and a simple ring.\nShe holds a {argument name="drink" default="clear plastic cup filled with a pale green iced matcha drink"} near her mouth with both hands, as if about to take a sip. The cup has a simple handwritten-style café label.\nBackground features lush mature green trees, a quiet paved walkway, elegant stone-and-metal fencing, and soft architectural details partially visible through the foliage. Bright natural daylight filters through the trees, creating beautiful dappled sunlight and soft shadows across the scene.\nKorean-inspired minimalist fashion photography, candid street-style aesthetic, soft summer atmosphere, natural daylight, realistic skin texture, detailed hair strands, realistic fabric and handbag textures, subtle warm color grading, shallow depth of field, gentle background bokeh, authentic smartphone photography, photorealistic, high resolution, natural body proportions, vertical 4:5 composition.\n cartoon, anime, CGI, plastic skin, excessive makeup, distorted face, asymmetrical eyes, malformed hands, extra fingers, distorted drink cup, warped handbag, unrealistic hair, unnatural body proportions, oversaturated colors, harsh shadows, blurry subject, excessive retouching, text artifacts, watermark.
使用此提示词 ↗
Product Marketing - Cinematic Luxury Fashion Portrait — output

Product Marketing - Cinematic Luxury Fashion Portrait

HeisenLegacyGPT Image 2
Ultra-realistic cinematic luxury fashion photograph of a {argument name="subject" default="strikingly handsome young man in his early 20s"} relaxing confidently in a {argument name="setting" default="pristine minimalist white bedroom"}. He has thick naturally wavy jet-black hair, strong well-defined eyebrows, deep expressive dark-brown eyes, a straight sculpted nose, prominent cheekbones, a sharp masculine jawline, and subtle natural stubble. He wears an {argument name="clothing" default="effortlessly elegant ivory knit sweater with tailored cream trousers"}, creating a refined monochromatic old-money aesthetic.\n\nThe bedroom features crisp white walls, soft white linen bedding, minimalist furniture, sheer floor-to-ceiling curtains, and warm natural sunlight gently streaming through the curtains, creating delicate patterns of light and shadow across the room. He stands beside the bed near the window, one hand casually resting on the curtain, gazing calmly toward the sunlight with a peaceful, introspective expression.\n\nSoft morning atmosphere, airy composition, warm highlights, subtle shadows, natural skin texture, realistic fabric details, sophisticated editorial styling, luxury lifestyle magazine aesthetic, cinematic depth of field, 85mm lens, f/1.8, soft bokeh, photorealistic, HDR, ultra-detailed, 8K, elegant neutral tones, serene and timeless mood, no artificial-looking skin, no exaggerated features, no text, no watermark.
使用此提示词 ↗
Product Marketing - European Autumn Fashion Editorial — output
Comic / Storyboard - The Last Train Home cinematic scene — output

Comic / Storyboard - The Last Train Home cinematic scene

Compound Learning - Tech | ScienceGPT Image 2
{
  "title": "The Last Train Home",
  "image_description": "A breathtaking cinematic scene at an {argument name=\"location\" default=\"abandoned train station lost in the mountains\"}. Thick fog rolls across the empty platform. Vintage lamps cast warm golden light through the mist. A lone traveler wearing a dark wool coat stands beside an old leather suitcase. An ancient train slowly emerges from the fog. Every window glows softly. Snow falls silently around the station. Above the platform hangs a weathered sign that reads: HOME. The traveler's eyes fill with emotion as he realizes the train should not exist.",
  "story_element": {
    "twist": "The train only appears once, exactly one hour before a person dies."
  },
  "environment": {
    "location": "Forgotten mountain railway station",
    "weather": "Snowfall and dense fog",
    "time": "Midnight"
  },
  "cinematography": {
    "shot_type": "Wide cinematic shot",
    "camera": "ARRI Alexa 65",
    "lens": "50mm anamorphic",
    "composition": "Symmetrical storytelling frame"
  },
  "lighting": {
    "type": "Warm vintage lamps mixed with cold moonlight",
    "effects": [
      "Volumetric fog",
      "Snow particles",
      "Atmospheric glow"
    ]
  },
  "mood": [
    "Emotional",
    "Mysterious",
    "Bittersweet",
    "Nostalgic"
  ],
  "quality": {
    "style": "Award-winning cinematic masterpiece",
    "realism": "Ultra photorealistic",
    "resolution": "8K",
    "IMAX_quality": true,
    "film_grain": true
  }
}
使用此提示词 ↗
Social Media Post - Spontaneous coastal road travel photo — output

Social Media Post - Spontaneous coastal road travel photo

NicoleGPT Image 2
A highly photorealistic candid photograph of a completely new {argument name="subject" default="young adult Asian woman"} standing beside a {argument name="location" default="scenic coastal road"} at {argument name="time of day" default="sunset"}, photographed from inside a parked car through the passenger-side window. She has long, naturally messy dark-brown hair being moved by the evening breeze, a calm slightly distant expression, and wears a dark oversized jacket over a simple casual top. She stands outside the car near a metal roadside guardrail, her body turned slightly away from the camera while she glances back toward the person photographing her. The camera is positioned low and inside the vehicle at an oblique side angle, with the curved car window frame, interior door, and side mirror partially visible in the foreground. The framing is imperfect and feels like a spontaneous travel photograph rather than a planned portrait. Behind her is a completely different coastal landscape with steep green hills, a winding road, scattered houses, rocky shoreline, and calm ocean water stretching toward the horizon. The sun is low near the horizon, creating soft golden highlights along her hair and the landscape while much of the car interior remains dark. Natural smartphone/compact digital-camera photography, realistic lens distortion and perspective, subtle handheld imperfections, authentic skin texture, individual flyaway hairs, realistic clothing fibers and folds, natural facial asymmetry, slight atmospheric haze, subdued contrast, gentle sunset colors, mild digital noise, very subtle film grain, realistic dynamic range, slightly imperfect exposure, soft background blur, no studio lighting, no beauty-filter skin, no excessive sharpening, no artificial CGI appearance. The image should feel like a genuine photograph taken spontaneously during a road trip. Highly photorealistic, vertical 3:4 composition.
使用此提示词 ↗
Social Media Post - Candid car selfie in sunlight — output

Social Media Post - Candid car selfie in sunlight

AqsaGPT Image 2
Ultra-realistic candid selfie of a {argument name="subject" default="young woman"} sitting inside a {argument name="location" default="car"}, captured from a slightly low front-facing smartphone camera angle. She has long, straight, silky dark-brown hair falling naturally over both shoulders, soft wispy strands around her face, delicate youthful features, fair natural skin, dark brown eyes, subtle makeup, and a cute playful closed-mouth smile. She is wearing an oversized white graphic T-shirt and a delicate silver bracelet. One arm is raised toward the camera, forming a playful hand gesture around one eye using her thumb and index finger, while the other fingers extend upward. She slightly squints one eye through the hand gesture and smiles naturally, creating a fun, carefree expression. Bright direct sunlight enters through the car window, illuminating her face and creating strong golden highlights across her hair and skin. Dark car interior surrounds her, with black leather seats and window frames visible. Through the open side window, a vivid blue sky with soft white clouds is visible. Natural sunlight and deep interior shadows create strong contrast. Authentic smartphone photography, casual social-media selfie aesthetic, natural skin texture, realistic hair strands, realistic hands and fingers, warm sunlight, high dynamic range, subtle lens imperfections, candid youthful atmosphere, photorealistic, highly detailed, sharp face, natural proportions, vertical 4:5 composition.
使用此提示词 ↗
Comic / Storyboard - Museum of Lost Tomorrows conceptual scene — output

Comic / Storyboard - Museum of Lost Tomorrows conceptual scene

Compound Learning - Tech | ScienceGPT Image 2
{
  "title": "The Museum of Lost Tomorrows",
  "image_description": "A vast {argument name=\"location\" default=\"underground museum hidden beneath a forgotten city\"}. The architecture is breathtaking, with endless marble hallways illuminated by warm golden light filtering through giant glass ceilings. Thousands of glass display cases stretch endlessly into the distance. Inside each case is a future that never happened. One contains a wedding dress that was never worn. Another holds a Nobel Prize medal that was never earned. Another displays a dusty guitar beside a sold-out world tour poster from a musician who gave up too soon. Visitors walk silently through the museum, staring at the exhibits with tears in their eyes.",
  "main_subject": {
    "character": "A young man standing alone",
    "appearance": "Dark overcoat, slightly worn shoes, tired eyes",
    "expression": "Shocked and emotional",
    "focus": "Sharp focus"
  },
  "story_element": {
    "twist": "At the end of the museum, he discovers an entire room dedicated to his own abandoned dreams.",
    "question": "What would you do if you could see the life you almost lived?"
  },
  "environment": {
    "location": "Underground Museum",
    "lighting": "Golden cinematic lighting",
    "atmosphere": "Silent, emotional, sacred"
  },
  "cinematography": {
    "camera": "ARRI Alexa 65",
    "lens": "50mm anamorphic",
    "shot_type": "Wide cinematic shot",
    "depth_of_field": "Shallow",
    "composition": "Award-winning storytelling frame"
  },
  "mood": [
    "Heartbreaking",
    "Thought-provoking",
    "Melancholic",
    "Beautiful"
  ],
  "style": {
    "reference": "Christopher Nolan meets Denis Villeneuve",
    "quality": "Ultra photorealistic",
    "resolution": "8K",
    "film_grain": true,
    "volumetric_lighting": true
  }
}
使用此提示词 ↗
Product Marketing - Miniature city in coffee cup — output
Comic / Storyboard - Cloud Friend Scrapbook Poster — output

Comic / Storyboard - Cloud Friend Scrapbook Poster

小小东GPT Image 2
Goal: Create a whimsical two-part aesthetic poster about a cute cloud character, combining a real-sky photo scene on top with a hand-drawn scrapbook guide below.

Canvas: Vertical 3:4 poster. The top half is a realistic outdoor sky photograph; the bottom half is a cream paper illustration panel with a soft grainy texture.

Top scene: A bright blue daytime sky with a few small white clouds. In the center, one fluffy white cloud is turned into a friendly character using simple black graphic features: exactly 2 dot eyes, 1 curved smiling mouth, and 2 raised bone-like arms angled upward from the left and right sides. The mood is light, airy, relaxed, and childlike. Use natural sunlight and keep the sky spacious.

Bottom panel title: Handwritten title centered near the top: {argument name="title text" default="Cloud Friend"}. Add a small blue star to the left, a red heart to the right, tiny blue sparkles, and a wavy blue underline beneath the title.

Bottom layout: Arrange exactly 9 small illustrated vignettes in a loose 3-by-3 grid, each with a handwritten numbered caption underneath. The style should look like colored pencil, crayon, and ink on warm off-white paper.

Nine vignettes and captions:
1. “hello!” — a small smiling cloud character inside a rough blue scribble patch, with two black arms and a tiny red heart.
2. “happy face” — a simple face made of 2 black dot eyes, 1 black smile, and 2 soft pink blush circles.
3. “arms up” — two separate black bone-like arms raised like a cheering gesture, with small blue motion marks and a red doodle accent.
4. “very blue” — a rounded rectangle of textured blue crayon sky containing exactly 2 small white clouds.
5. “soft & fluffy” — a pale gray pencil sketch of one fluffy cloud outline, with a small red heart.
6. “warm sun” — a yellow-orange circular sun with short radiating rays.
7. “gentle breeze” — light blue flowing wind lines with a small swirl and red doodle accent.
8. “daydream view” — a tiny simple landscape with exactly 2 green trees, a green ground line, and a small red-roof house outline.
9. “good mood” — a cheerful doodle cluster containing 1 blue star, 1 pink heart, 1 mint green cloud, small dots, and red accent lines.

Footer text: Centered along the bottom in small handwritten lettering: {argument name="footer quote" default="Look up. Little joys are floating around."} followed by a small red heart and blue dots.

Visual style: Mix realistic photography in the top half with naive hand-drawn journaling in the bottom half. Use soft blue, cream, black, pastel pink, yellow, mint, and red accents. Keep all text handwritten, imperfect, casual, and charming.

Constraints: Preserve the exact two-panel composition, the exact count of 9 numbered vignettes, and the friendly minimalist cloud face. Avoid extra characters, logos, watermarks, or dense text.
使用此提示词 ↗
Game Asset - Stylized Pine Tree Game Asset Sheet — output

Game Asset - Stylized Pine Tree Game Asset Sheet

OrcDevGPT Image 2
Goal: Create a clean game-asset concept sheet for a stylized conifer tree labeled {argument name="asset code" default="TREE01"}, designed for conversion into a 3D model.

Canvas: Wide horizontal concept-art sheet on a plain light gray background, soft studio lighting, no environment, no decorative border. Use muted natural colors and a painterly semi-realistic fantasy game style.

Layout: Show exactly 4 tree depictions: 1 large front view on the left, 1 large side view in the center, 1 large back view on the right, and 1 smaller game-scale view near the bottom center. Add exactly 2 scale references: a vertical height ruler beside the large front tree labeled {argument name="tree height" default="9 m"}, and a gray human silhouette beside the small game view with a short vertical ruler labeled {argument name="human height" default="1.8 m"}. Place view labels under the trees: {argument name="front label" default="FRONT"}, "SIDE", "BACK", and "GAME VIEW".

Subject details: The tree is a tall spruce or fir with a rugged straight trunk, visible gray-brown cracked bark, exposed root flares at the base, and layered drooping branch tiers. The foliage is dense but irregular, made of jagged clustered needle masses in desaturated olive green, dark pine green, and shadowed charcoal green. The silhouette is triangular and natural, with a pointed crown, asymmetric branch lengths, some bare branch sections visible near the trunk, and broad lower boughs tapering upward. The side view should look narrower with overlapping branches, while the back view should reveal similar foliage massing and trunk structure from behind. The small game view should repeat the same tree design at reduced scale, next to the human silhouette for proportion.

Typography and annotations: Put the asset code in the upper-left corner in a simple serif typeface, large and black. Use small black serif labels beneath each view. Rulers are thin dark gray vertical lines with small caps and understated measurement text.

Constraints: Keep the sheet uncluttered and orthographic, with consistent scale among the three large views. Use exactly the 4 tree depictions and 2 scale references described, avoid extra plants, rocks, UI elements, textures, logos, watermarks, or background scenery.
使用此提示词 ↗
Social Media Post - Wimbledon Motion Diptych — output

Social Media Post - Wimbledon Motion Diptych

小小东GPT Image 2
Goal: Create a two-panel editorial sports artwork about {argument name="tennis player" default="Roger Federer"} at Wimbledon, combining a rough painterly match scene with a minimalist poster illustration.

Canvas: Wide horizontal 16:9 composition split exactly into 2 equal vertical panels. Left panel is a dense, realistic, rough-brush illustration of a grass-court tennis match; right panel is a clean off-white poster with distressed typography and collage-like tennis illustration. No outer border.

Layout: Left panel occupies the left half and shows an action moment from behind the near player, lunging low across a worn green grass court toward the net. The far baseline has 1 opposing player in white preparing to return, with 2 ball kids or officials standing near the back wall. A large scoreboard sits high at center-left behind the court, with a packed audience filling the stands above it. Right panel occupies the right half with lots of negative space: large hand-painted green title at upper center-right, small slogan text at mid-left, a running tennis player illustration across the lower center, and a small vertical text block near the lower right.

Left panel subject details: Depict exactly 2 tennis players: 1 foreground player in all-white tennis outfit, white headband, white shoes with small red accents, stretched into a dramatic forehand reach with racket extended; 1 distant opponent in white near the far baseline. Include 1 yellow tennis ball visible near the far player. The court is Wimbledon-style grass with scuffed, patchy texture, white boundary lines, a black net with yellow posts, and dark green walls. Use a crowded stadium background with many tiny spectators in hats and summer clothing. The left panel should feel like a photo transformed into thick oil-paint strokes, with visible palette-knife texture and rough edges.

Scoreboard text: Include a black-and-green Rolex-style scoreboard with yellow lettering. It should show the names Marin CILIC and Roger FEDERER, with Federer leading; include readable fragments such as "ROLEX", "Marin CILIC", "Roger FEDERER", "PREVIOUS SETS", "SETS", "GAMES", "POINTS", and small challenge remaining lines, but do not worry about perfect numerical accuracy.

Right panel visual style: Minimalist off-white paper background with subtle grain. At the top, write {argument name="headline text" default="Still in Motion"} in large dark forest-green, rough dry-brush, slanted hand-lettering. The illustrated player below should match the foreground player from the left panel, drawn as a low-poly paper collage / torn-paper texture in white, tan, brown, gray, and green, sprinting to the right with a black racket and a small yellow tennis ball floating ahead. Add exactly 3 rough green brush-stroke court marks beneath the player.

Right panel text content: Add two small serif text blocks: on the left of the illustrated player, stacked lines reading {argument name="left slogan" default="Same Game A Deeper You"} with a short horizontal rule beneath; near the lower right, stacked lines reading {argument name="right slogan" default="TENNIS LIVES DIFFERENTLY HERE."} with a short horizontal rule beneath.

Color and mood: Wimbledon greens, white clothing, sunlit grass, dark scoreboard, warm crowd tones, and an elegant vintage sports-magazine mood. The overall finish should be rough, textured, slightly imperfect, artistic, and editorial rather than glossy.

Constraints: Use exactly 2 panels, exactly 2 tennis players, exactly 1 visible tennis ball in the left panel, exactly 1 visible tennis ball in the right panel, exactly 1 large scoreboard, exactly 3 green brush-stroke court marks in the right panel. Avoid logos other than the scoreboard-style Rolex reference, avoid watermarks, and keep the right panel spacious and uncluttered.
使用此提示词 ↗
Comic / Storyboard - Undead Vampire King Throne — output

Comic / Storyboard - Undead Vampire King Throne

Lee PenkmanGPT Image 2
Create an ultra-detailed dark fantasy gothic illustration of {argument name="character name" default="an undead vampire king"} enthroned inside a ruined cathedral throne room. The central figure is a pale, corpse-like monarch with long flowing {argument name="hair color" default="white hair"}, glowing red eyes, blood running from the mouth, and a jagged black crown. He wears ornate blackened armor covered in skulls, spikes, chains, red jewels, and torn crimson-black royal robes. He sits arrogantly on a skull-encrusted stone throne, one hand resting on the armrest and the other gripping a long ceremonial sword with a skull pommel, gothic crossguard, red gem, and pale blade. Surround him with exactly 9 clearly visible supernatural figures or creatures: 1 large translucent ghostly warrior hovering behind him with a sword, 1 smaller floating ghost near the moonlit archway, 4 misty human spirits on the stairs and balcony in the midground, 2 armored spectral soldiers at the lower right holding weapons, and 1 snarling ghost wolf in the lower left foreground. Set the scene in a vast black stone cathedral with broken pointed arches, towering columns, candles, chains, cracked steps, skulls, and tattered red banners. Include exactly 3 readable banner or stone inscriptions: left hanging banner says {argument name="left banner text" default="BLOOD BEYOND DEATH"}, right hanging banner says {argument name="right banner text" default="KINGS DIE BUT LEGENDS HAUNT FOREVER"}, and the stone block near the lower left says {argument name="stone inscription" default="SWORDS REMEMBER WHAT MEN FORGET"}. Through the huge archway, show a distant spired gothic castle, an arched bridge, flying bats, storm clouds, cold mist, and a huge crimson blood moon. Use cinematic low-angle composition, hyperreal digital painting, intricate concept-art detail, cold blue-gray fog contrasted with deep crimson accents, dramatic moonlight, candle glow, volumetric mist, sharp armor highlights, macabre regal atmosphere, no modern objects, no cartoon style.
使用此提示词 ↗
Game Asset - Futuristic Sky Runner Robot — output

Game Asset - Futuristic Sky Runner Robot

野生の男GPT Image 2
Create a vibrant 3D stylized endless-runner game concept image in a wide 16:9 cinematic frame. Show one cute humanoid robot mascot mid-jump from behind and slightly to the side, running across floating futuristic platforms high above the clouds. The robot has a glossy white rounded helmet with a black face screen and cyan glowing eyes, small antenna, segmented cream-and-gray limbs, blue gloves and oversized blue boots, plus a flowing {argument name="scarf color" default="golden yellow"} scarf trailing dramatically in the wind. The world is a cheerful futuristic {argument name="city setting" default="Tokyo-inspired sky city"} at {argument name="time of day" default="warm sunrise"}, with soft golden light, bright blue sky, stylized puffy clouds, and a distant skyline. Include one snow-capped {argument name="background landmark" default="Mount Fuji"} on the left horizon, one sleek elevated bullet train curving along a rail on the left, and one large red-and-white torii gate platform on the right. The runner path is made of clean cream-colored floating platforms with turquoise-blue tile panels, rounded edges, and colorful support pillars emerging from clouds. Place exactly seven floating collectible crystal gems ahead of the robot along the path: one large pink diamond crystal near the center, one orange crystal, one blue crystal, one magenta crystal, one small pale yellow crystal, one small orange crystal, and one tiny cyan crystal farther into the distance. Use a polished Pixar-like game art style, soft global illumination, saturated colors, crisp toy-like materials, dynamic perspective, strong sense of motion, and no text, UI, logos, or watermarks.
使用此提示词 ↗
Game Asset - Galaxy Rail Title Screen — output

Game Asset - Galaxy Rail Title Screen

⛩🧙‍♂️ゲーム仙人かいぽん⚔️ぱぱら快刀🌻💙GPT Image 2
Goal: Create a retro 2D vertical arcade game title screen for a rail shooter called {argument name="game title" default="GALAXY RAIL"}, with a space-and-railroad theme and Japanese UI text.

Canvas: Portrait mobile game splash screen, 9:16 aspect ratio, pixel-art inspired but clean vector-like illustration, dark navy outer-space background filled with many small white and pale yellow stars.

Layout: Center the title in the upper-middle. Place a large partially cropped Earth-like blue planet on the left behind the title, occupying about one third of the image height, with simple green landmasses and translucent cloud streaks. Add exactly 3 vertical railroad tracks running from top to bottom: one left track partly over the planet, one center track, and one right track; each track has gray rails and many brown sleepers. Put exactly 2 small train/player cars on the lower center track, stacked vertically with rounded white bodies, blue windows, yellow outlines, and a yellow headlight on the top car.

Text content: Main title text is two lines: “GALAXY” in large white slab-serif letters with a dark navy outline, and “RAIL” below in large golden yellow slab-serif letters with a dark navy outline. Under the title, show the Japanese tagline {argument name="tagline" default="銀河を、レールで切り拓く"} in bright cyan. Beneath it, add small gray feature text: “RAIL SHOOTER / 2D VERTICAL / 3 STAGES”. In the lower third, add a translucent dark blue rectangular instruction panel with a thin blue border. Inside the panel, show three Japanese instruction lines: “←→↑↓ / WASD:移動(レールから自動ジャンプ)”, “Z / SPACE:短押し=単発 / 長押し=2連射”, and “時間経過で最寄りのレールへ強制着地”. Below the panel, add a small red warning line: “赤いサークル=瞬着攻撃の予兆。収束前に発生源を倒せ!” At the very bottom, centered, show pale gray prompt text: {argument name="start prompt" default="PRESS Z SPACE OR CLICK TO START"}.

Visual style: Retro arcade title screen, simple 2D game mockup, crisp shapes, slightly pixelated typography, high contrast, deep-space palette of navy, white, cyan, yellow, gray, and red. The composition should feel like a vertical rail shooter prototype splash screen.

Constraints: Use exactly 3 full-height vertical tracks, exactly 2 train cars, one planet on the left, one instruction panel, and no characters or extra UI windows. Keep all text legible and centered, with no logos or watermarks.
使用此提示词 ↗
Social Media Post - Smiling Bento Shop Sign Portrait — output

Social Media Post - Smiling Bento Shop Sign Portrait

vevrier90333GPT Image 2
Create a highly realistic vertical smartphone photo of a smiling young adult Japanese woman standing outside a busy bento shop in bright midday sunlight. She has a short glossy black bob haircut, soft makeup, rosy cheeks, pink lipstick, and a friendly direct gaze toward the camera. She wears a frilly pastel pink halter apron-style dress with ruffled straps and a bow at the waist. She holds a translucent frosted acrylic sign in both hands at chest level, with visible screw holes in the corners and handwritten marker art on the front. The sign contains exactly 4 main drawn elements: 1 pink name text reading {argument name="character name text" default="ANRIさん"}, 1 blue encouragement phrase reading {argument name="message text" default="今日も頑張ってください!"}, 2 pink heart doodles, and 1 small colorful cartoon bento box illustration with rice, side dishes, and vegetables. Background: storefront glass windows reflecting the street, shelves and product displays inside, Japanese bento posters and menu boards around the entrance, including a vertical pink-and-white banner on the left with large black Japanese characters for bento. Use shallow depth of field but keep the woman and sign crisp. Lighting should be natural, sunlit, cheerful, and slightly glossy, with realistic skin texture, reflections on the acrylic board, and candid social-media-photo framing. Do not add extra people, extra signs in the foreground, watermarks, or unrelated text.
使用此提示词 ↗
Comic / Storyboard - Anime Train at Dusk with Mount Fuji — output

Comic / Storyboard - Anime Train at Dusk with Mount Fuji

ChiyoGPT Image 2
Create a cinematic anime-style illustration of a quiet train interior at dusk, with a melancholic, dreamlike atmosphere. The scene is viewed from inside an almost empty commuter train, looking across a long dark blue bench seat and large rectangular windows. A single petite teenage girl sits near the right side of the bench, hands folded in her lap, gazing slightly to the side with a calm, lonely expression. She wears an all-black gothic-inspired outfit: black beret, dark blouse with ribbon details, long black skirt, dark socks, and black lace-up boots; her {argument name="hair color" default="black"} hair is styled in two long braids, and she has pale skin and delicate anime facial features. Through the train windows, show a serene Japanese lakeside landscape with {argument name="mountain" default="Mount Fuji"} centered in the distance, reflected softly in the water, with small pagoda silhouettes and treetops along the shore. The sky outside is pastel twilight with lavender-blue tones, pink clouds, and one large round {argument name="sun or moon" default="pink setting sun"} glowing near the upper left window; include a few pink cherry blossom branches and petals near the top left. Inside the train, keep the lighting very dim and cool, with the girl mostly in shadow against the glowing scenery outside. Include exactly three visible hanging train straps near the ceiling, one large multi-pane window set behind the bench, and one train door with tall narrow windows on the right. Use a wide 16:9 composition, soft painterly cel-shaded anime rendering, muted navy interior shadows, pastel pink and blue exterior light, no text, no watermark, no extra passengers.
使用此提示词 ↗
Product Marketing - Woman With Matching 3D Bobblehead — output

Product Marketing - Woman With Matching 3D Bobblehead

Laraib Fatima‎GPT Image 2
Create a high-resolution, ultra-realistic studio composition in a 4:5 vertical aspect ratio showing exactly two subjects: one real young woman and one 3D bobblehead caricature version of the same woman standing beside her. The real woman is tall and positioned on the right, with long slightly messy dark brown wavy hair, wispy bangs, round thin-rimmed eyeglasses, a calm but mildly scolding expression, and a casual confident pose. She wears a light blue short-sleeve polo shirt with a small button placket, high-waisted loose white wide-leg pants, white socks, and white sneakers. Her left hand rests at her hip while holding a plain white baseball cap, and her right arm bends forward with her index finger pointing down toward the caricature. The caricature stands on the left at about one-third the real woman’s height, with an oversized head, tiny body, large glossy expressive eyes behind matching round glasses, the same long dark wavy hair and bangs, a cute mischievous smile, and a slightly bashful pose with both hands tucked behind her back. Dress the caricature in the same outfit: light blue polo shirt, white high-waisted pants, and white sneakers. Use {argument name="character appearance" default="a young East Asian woman with long dark brown wavy hair, wispy bangs, and round glasses"}, {argument name="shirt color" default="light blue"}, {argument name="pants color" default="white"}, {argument name="real woman pose" default="standing with one hand on hip holding a white cap and the other hand pointing down"}, and {argument name="caricature expression" default="cute mischievous smile looking upward"}. Set the scene in a clean seamless gray studio with soft gradient background, smooth concrete floor, subtle shadows under both figures, realistic photographic lighting, crisp fabric detail, natural skin texture, and a polished premium editorial look. Make the bobblehead stylized but physically present and rendered as a high-quality 3D figure, while the real woman remains photorealistic. No text, no logos, no extra people, no watermark.
使用此提示词 ↗
Infographic / Edu Visual - Golden Deer Morning Felt Transformation — output

Infographic / Edu Visual - Golden Deer Morning Felt Transformation

小小东GPT Image 2
Goal: Create a two-panel horizontal comparison image showing a real-life nature photo transformed into a minimalist needle-felt illustration.

Canvas: Wide 16:9 canvas split vertically into two equal halves with a clean straight division down the center.

Left panel: A realistic golden-hour forest meadow photograph. Show exactly 5 deer in the foreground: 1 deer walking at far left, 1 deer grazing left-center, 1 deer grazing center, 1 small deer standing upright center-right, and 1 larger deer standing at far right. The meadow is covered in dewy green grass, with warm yellow sunlight and soft morning mist. In the background, dark evergreen trees form a dense forest silhouette, with additional leafy branches at the upper right. Strong sunbeams stream diagonally through the trees from the upper right, creating a serene cinematic atmosphere.

Right panel: A soft handmade needle-felt wool reinterpretation of the same scene on a warm off-white textured paper background with generous blank space. Center the felt scene horizontally in the upper-middle area. Show exactly 5 small felt deer matching the left panel arrangement: 1 standing deer at far left, 1 grazing deer left-center, 1 orange-brown grazing deer center, 1 small upright deer center-right, and 1 larger standing deer at far right. Place them on a thin strip of fuzzy pale green felt grass. Above them, include a simple round felt sun and 2 soft horizontal felt cloud wisps. Under the felt grass, add handwritten text reading {argument name="caption text" default="A kinder morning ♡"} in a delicate gray-brown script.

Visual style: Combine warm healing aesthetics, generous white space, cozy wool-felt texture, soft edges, muted earthy colors, and a calm poetic mood. The left side should look photographic and atmospheric; the right side should look like a handcrafted felt illustration with visible fibers and a minimal composition.

Constraints: Keep the split-screen layout, exactly 5 deer on each side, no extra animals, no border, no watermark, no additional text beyond the caption.
使用此提示词 ↗
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 ↗