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

Social Media Post - Tropical Vacation Editorial Photography — output

Social Media Post - Tropical Vacation Editorial Photography

𝑨𝒇𝒓𝒊𝒏GPT Image 2
Create a vibrant, cinematic, photorealistic image inspired by the provided reference image. Use the reference only for the overall character, mood, composition, and fashion aesthetic — create a fresh original scene.

A beautiful young woman with long, naturally wavy dark-brown hair, warm expressive eyes, glowing natural skin, and a bright genuine smile. She has a relaxed, carefree personality and looks naturally happy. Her hair flows dramatically in the warm ocean breeze.

She wears a loose, flowing {argument name="dress color" default="coral-red/orange"} oversized dress with {argument name="pattern" default="large cream botanical and floral patterns"}, lightweight fabric, wide sleeves, and a relaxed vacation silhouette. The dress moves naturally with the wind, creating elegant flowing shapes.

Place her in a {argument name="setting" default="stunning tropical coastal environment with turquoise ocean water, soft waves, golden sandy beach, lush green tropical plants, distant mountains, palm trees, and colorful flowers"}. Use rich, vibrant colors with a warm cinematic golden-hour atmosphere.

Photography style: ultra-realistic editorial travel photography, cinematic composition, natural skin texture, realistic facial details, subtle film grain, soft depth of field, beautiful bokeh, volumetric sunlight, warm highlights, rich color grading, high dynamic range, realistic fabric texture, natural hair movement.

Composition: full-body portrait, slightly dynamic walking pose, body turned naturally toward the camera, looking directly into the lens with a joyful smile. Capture a candid vacation moment rather than a posed studio portrait.

Make the image feel colorful, dreamy, youthful, spontaneous, luxurious, and cinematic while remaining photorealistic.

Avoid: distorted anatomy, extra fingers, extra limbs, duplicated body parts, malformed hands, unnatural face, plastic skin, excessive makeup, stiff posing, warped clothing patterns, floating objects, artificial-looking hair, oversaturated skin, blurry face, low detail, text, watermark, logo.

Vertical portrait composition, 9:16 aspect ratio, extremely detailed, photorealistic, cinematic, vibrant tropical color palette.
使用此提示词 ↗
Comic / Storyboard - Spectral Courier Gothic Alley — output
Product Marketing - Pastel Flower Petal Dress — output

Product Marketing - Pastel Flower Petal Dress

Prompt アトリエ|AI画像プロンプトGPT Image 2
Subject:
{argument name="main theme" default="Afternoon wrapped in flower clothing"}

Main Subject:
In the center of the frame, a woman in a dress made only of pink, white, and pale purple small flowers sits sideways on a floor covered in petals. Behind her, a giant wall of the same-colored flowers extends vertically, and she holds a flower twig next to her face with her right hand.

Person/Expression:
The woman turns her face to the front, looks at the camera, and smiles modestly. A round-cheeked oval face with a small chin, horizontally long bright eyes, thin loose eyebrows, a small nose tip, and pale coral lips. {argument name="hair style" default="Dark brown hair"} is in a high braided updo, with thin bangs, curled wisps on both sides, and small flowers in the hair. Pale pink eyes and cheeks, smooth bright skin.

Clothing/Pose:
A sleeveless floral dress densely packed with small roses, star-shaped flowers, and clustered pale purple flowers. It features a deep V-neck, shoulder straps covered in flowers, a waist that follows the body, and a long floral train flowing from the left hem to the floor. Resting her left hand on the floor for support, bending both legs toward the right of the screen, and holding a flower branch to her face with her right hand.

Background/Light:
A flower wall of pink, white, and pale purple from the center to the right, a pale beige wall and floor space on the left, and a large amount of small flowers and petals in the foreground. {argument name="light softness" default="Soft direct light"} from the upper left illuminates the flowers and skin, scattering small flower shadows on the wall.

Composition/Camera:
3:4 vertical orientation, slightly high camera angle capturing from the head to above the knees. The person is large in the center, and the long flower train spreads to the lower left foreground. Focus on the person and the foreground flowers, slightly blurring the background flower wall.

Texture/Style:
A photorealistic concept photo that finely depicts realistic petals and skin. Summarizes pink, white, and pale purple in a soft pastel tone.

Negative:
Do not make the dress made only of flowers out of fabric; do not omit the long floral train
使用此提示词 ↗
Profile / Avatar - South Asian Portrait with Majestic Peacock — output
Profile / Avatar - Ultra High-Res Photo Restoration — output
E-commerce Main Image - Luxury Serum Splash Advertisement — output
Product Marketing - RAKIB Social Media Design Poster — output

Product Marketing - RAKIB Social Media Design Poster

Rakib MahmudGPT Image 2
Create a vertical social media promotional poster for {argument name="brand name" default="RAKIB"}, advertising {argument name="service line" default="Social Media Designs"}. Use a deep emerald-green studio background with soft smoky abstract waves and a subtle glowing aura behind the headline. Center a smiling young Black male designer/marketer seated in a modern gray armchair, wearing a pale sage-green button-up shirt and beige pants, holding a dark tablet with both hands. Make the portrait realistic, polished, and commercial, with soft front lighting and a confident friendly expression.

Canvas and layout: portrait 2:3 ratio, poster-style composition. At the very top, place the huge uppercase word “RAKIB” in bold sans-serif letters, glowing pale mint-white. Directly below it, place “Social Media Designs” in an elegant white serif font. The person occupies the center lower half, overlapping the floating content cards around him.

Floating cards: include exactly 4 tilted Instagram-style post mockup cards surrounding the subject, each with a white/cream background, rounded corners, tiny social icons, and the small header name “haris designs” with a green circular logo. Card 1 at upper left: headline “Grow Your Brand With AI,” with plant imagery and a green call-to-action pill. Card 2 at upper right: headline “Creative Content That Converts,” with a circular green interior/product image. Card 3 at lower left: headline “Design Smarter Not Harder,” with a leaf image. Card 4 at lower right: headline “Stand Out Stay Ahead Go Viral,” with a plant image. Keep all card text small but readable enough to imply social media design templates.

Bottom feature strip: add a dark translucent green bar across the bottom with exactly 4 feature groups separated by thin vertical divider lines. From left to right: an AI/star icon labeled “AI-Powered Creativity”; a lightbulb icon labeled “Eye-Catching Designs”; a chat bubble icon labeled “Engaging Content”; an upward bar chart icon labeled “Grow Your Brand.”

Visual style: premium digital marketing ad, realistic portrait combined with clean UI mockups, mint-green glow accents, high contrast, crisp typography, professional Instagram growth branding. No watermark, no extra logos beyond the mockup headers, no additional people, and no clutter.
使用此提示词 ↗
Profile / Avatar - Sloppy Doodle Cat Portrait — output
Product Marketing - Spanish iPhone Pro Landing Page — output

Product Marketing - Spanish iPhone Pro Landing Page

NixusGPT Image 2
Goal: Create a realistic browser-like screenshot of a dark Apple-style product landing page for {argument name="product name" default="iPhone 17 Pro"}, shown inside a charcoal comparison/viewer frame.

Canvas: Square 800×800 image with a very dark gray outer background. Add a thin neon-green rounded border around the whole viewer frame. In the top left of the frame place small green monospace text {argument name="viewer label" default="lina-f-alpha"}. In the top right place two small controls: a download icon and a rounded dark button labeled “Edit” with a pencil icon.

Main webpage card: Center a large rounded-rectangle browser/page panel, approximately 660 px wide and 660 px tall, with clipped corners. The page is black with a dark top navigation bar.

Navigation: Include exactly 9 navigation items across the top bar, small light-gray text: “Mac”, “iPad”, “iPhone”, “Watch”, “Vision”, “AirPods”, “TV y Casa”, “Entretenimiento”, “Soporte”.

Hero section: Use a premium Apple keynote aesthetic. At the top center of the black hero area show the product name “iPhone 17 Pro” in white. Below it place a huge bold Spanish headline {argument name="headline text" default="Más Pro. Siempre."} with a soft cream-to-orange gradient. Under the headline add the subheading {argument name="subheading text" default="Un salto enorme en potencia, cámaras y autonomía."} in muted light gray. Below the subheading add exactly 2 pill buttons: a filled white button reading “Más información” and an outlined black button with white border reading “Comprar”.

Product image: The lower two-thirds of the page is dominated by an extreme close-up render of the back of a copper/orange iPhone, cropped so only the top back and camera island area are visible. Show exactly 3 large raised black camera lenses with copper metallic rings: one near the upper left, one lower left, and one near the upper middle. On the right side show exactly 3 smaller elements: a silver circular flash near the upper right, a tiny small dot below it, and a black circular sensor below-right. The phone body should have warm metallic reflections, subtle texture, rounded edges, and glossy shadows.

Bottom feature strip: Along the bottom of the webpage, overlay a dark translucent strip with exactly 4 feature cards separated by thin vertical divider lines. The leftmost card is partially cut off at the left edge, showing only the tail of text ending in “nites.” The second card has a white outline camera icon and text “Sistema de cámaras Pro” with subtext “Más detalle en cada historia.” The third card has a white battery icon and text “Autonomía para todo el día” with subtext “Llega más lejos.” The fourth card is partially visible on the right, with a white phone-outline icon and cropped Spanish text beginning “Pa…” and subtext beginning “Un…”.

Visual style: Photorealistic product render, high contrast, black luxury technology design, Spanish Apple-inspired typography, clean spacing, subtle glow on the copper finish, crisp UI elements, no watermark. Keep all text legible and do not add extra navigation items, extra buttons, or extra camera lenses.
使用此提示词 ↗
Profile / Avatar - Executive Office Phone Call Portrait — output

Profile / Avatar - Executive Office Phone Call Portrait

Abkr SadiqGPT Image 2
Create an ultra-realistic vertical professional lifestyle portrait of a {argument name="subject" default="young Black adult man"} seated confidently behind a polished wooden executive desk in a modern office, looking directly at the camera with a calm, focused, businesslike expression. He wears a crisp white dress shirt, a navy blue tie with tiny white dots, and a dark wristwatch, sitting in a tall black leather executive chair. His left hand holds a black corded telephone receiver to his ear, with the coiled cable leading off to the right, while his right hand writes with a black pen on a small yellow sticky note pad on the desk. The scene should feel like focus, confidence, and an important call in progress.

Use a realistic office background with a dark charcoal wall and a wooden bookshelf on the right containing exactly 4 black ring binders, a small potted plant, a row of orange-and-white books, and a black globe on an upper shelf. On the left wall, include a white crowned lion head logo above the brand text {argument name="brand text" default="FACELESS CONTENT"}. Add exactly 5 productivity notes on the wall: 4 colored sticky notes and 1 checklist sheet. The 4 sticky notes must read: yellow note “PLAN - FOCUS - EXECUTE”, blue note “DISCIPLINE TODAY SUCCESS TOMORROW”, pink note “BE BETTER THAN YESTERDAY”, and green note “HARD WORK BEATS TALENT”. The checklist sheet must be titled “DAILY CHECKLIST” and list exactly 5 checked items: “MEETING”, “CALLS”, “EMAILS”, “PLANNING”, and “FOLLOW UP”.

On the desk, include exactly 6 visible work items: a black mug printed with three stacked words {argument name="mug text" default="Discipline Focus Execution"}, a black calculator, a neat stack of multicolored sticky notes, the yellow note pad he is writing on, a clipboard document titled {argument name="document title" default="BUSINESS PLAN"} with simple charts, and an open silver laptop on the right edge. Use warm realistic lighting, sharp facial detail, natural skin texture, shallow-to-moderate depth of field, polished wood reflections, and a high-end corporate editorial photography style. Frame as a vertical 2:3 portrait, waist-up composition, subject centered, desk filling the lower third, no extra people, no surreal elements, no distorted hands, and no watermark.
使用此提示词 ↗
Profile / Avatar - Black-and-White Windowsill Portrait — output

Profile / Avatar - Black-and-White Windowsill Portrait

Abkr SadiqGPT Image 2
Create an ultra-realistic black-and-white lifestyle portrait of {argument name="subject" default="a young Black man"} sitting on the right side of a wide indoor windowsill beside {argument name="pet" default="a tabby cat"}, captured in a quiet, intimate moment with no text. The man has short close-cropped hair and a calm gentle expression, wearing {argument name="outfit" default="a plain white oversized T-shirt, dark denim shorts, white low-top sneakers, a wristwatch, and a ring"}. He sits leaning slightly back against the window frame with his arms crossed, legs extended downward and ankles loosely crossed, looking down toward the cat with a soft smile. The cat sits upright on the left side of the sill, facing the man and looking up at him. Use a vertical portrait composition, full body visible from head to shoes, with a large industrial-style window behind them divided by dark metal mullions; outside the glass, show softly blurred apartment buildings, balconies, trees, and bright daylight. The interior is minimal, with a deep white windowsill, plain wall, smooth floor, and gentle natural window light. Style the image as a candid documentary fashion photograph, high-resolution monochrome, realistic skin texture, soft contrast, shallow depth of field, peaceful atmosphere, quiet companionship, no props beyond the watch and ring, no watermark, no captions.
使用此提示词 ↗
Profile / Avatar - Watercolor Travel Portrait — output
Game Asset - Solemn Knight on Ashen Battlefield — output
Comic / Storyboard - Galaxy Train in an Old Classroom — output

Comic / Storyboard - Galaxy Train in an Old Classroom

春永睦月 Harunaga MutsukiGPT Image 2
Create a cinematic vertical anime illustration of a quiet old art classroom after rain, where {argument name="character name" default="a teenage schoolgirl"} sits sideways on the edge of a worn wooden desk near the left side of the room. She has {argument name="hair color" default="long wavy dark brown hair"}, soft expressive eyes, a calm wistful expression, and wears {argument name="outfit" default="a cream cardigan over a teal blouse with a long navy skirt and brown shoes"}. The classroom is warmly lit by late-afternoon sunlight pouring through exactly three large window sections on the right wall, with sheer white curtains glowing and drifting slightly; outside, show a wet school courtyard, greenery, and an old school building reflected in rain puddles. Include exactly one chalkboard on the back-left wall with faint chalk writing and pinned papers, exactly two wooden desks in the foreground and right side, one open sketchbook with architectural drawings on the front desk, jars of brushes, books, stacked papers, small framed paintings, a plaster bust, stools, and aged wooden floorboards. The key magical element is the girl's shadow stretching across the floor and transforming into {argument name="fantasy element" default="a luminous miniature galaxy train"}: a silver-white train with multiple connected cars curving through a star-filled cosmic river on the classroom floor, complete with glowing rails, constellations, tiny planets, nebula dust, sparkles, floating paper fragments, and delicate bridges or station-like structures. Blend realism and fantasy seamlessly so the cosmic train appears to emerge from the dark shadow beneath her feet. Use detailed Japanese anime painterly style, rich environmental storytelling, soft golden rim light, volumetric sunbeams, rain-afterglow atmosphere, intricate highlights, deep navy cosmic blues contrasted with warm amber classroom light, high detail, vertical composition, no watermark, no readable modern signage, no extra characters.
使用此提示词 ↗
Comic / Storyboard - Gold Two-Tone Anime Witch Poster — output

Comic / Storyboard - Gold Two-Tone Anime Witch Poster

おわ DaysGPT Image 2
Create a vertical 2:3 anime-style illustration in elegant two-color printing, using only pale warm cream paper and deep metallic {argument name="ink color" default="gold-brown"} ink, with no gradients except subtle ink-density variation. Show a graceful faceless witch-like young woman in profile facing right, her head slightly bowed, wearing an oversized wide-brim hat decorated with a ribbon band and exactly 5 small blossoms. She has very long flowing {argument name="hair color" default="gold-brown"} hair rendered as sweeping ribbon-like strands that arc dramatically leftward across the page, mixed with thin sketch lines and solid silhouette shapes. Her outfit is an ornate dark Victorian fantasy dress with an open back, corset-like waist, lace trim, layered skirt ruffles, long trailing fabric, and fine golden linework over large filled shadow areas. Behind her left side, add a large abstract butterfly-wing or ribbon flourish made of looping hair-like curves and translucent line shapes. Surround the figure with exactly 8 separate floral motifs: 3 large roses, 1 medium rose, 2 small five-petal flowers on the right, and 2 small lotus-like flowers near the lower left; also add about 18 scattered leaf-shaped petals. Add exactly 9 hanging geometric ornaments along the top and left edges, including diamonds, drops, long spear shapes, and one circular citrus-slice medallion near the top right, plus one black starburst ornament on the lower left. Composition should feel like a decorative art nouveau poster, high contrast, intricate golden line art, asymmetrical negative space, no visible eyes or facial features, no background scenery. Put a small handwritten signature reading {argument name="signature text" default="oura days 🐾"} in the lower-left corner.
使用此提示词 ↗
Profile / Avatar - Sweaty Anime Runner Portrait — output

Profile / Avatar - Sweaty Anime Runner Portrait

🐹🐹GPT Image 2
Create a polished anime-style vertical sports illustration of {argument name="character description" default="an androgynous young runner with pale skin, short white-silver hair tied into a small messy ponytail, and golden amber eyes"}. Show the character in close-up side profile facing right, cropped from the upper torso upward, with their bent arm raised near their mouth as if wiping sweat or catching their breath after running. They wear {argument name="running outfit" default="an orange racing singlet with black trim over a dark sports layer and a white race bib showing the number 4.8"}. The expression is calm, focused, slightly tired, with parted lips and a distant gaze. Add many visible sweat droplets on the face, neck, shoulder, arm, and hairline, catching bright sunlight. Use a vivid clear {argument name="background color" default="saturated blue sky"} as a simple flat background with no scenery. Lighting should be intense summer daylight from the upper right, creating glowing rim light on the hair and skin, crisp highlights, soft shadows, translucent skin tones, and delicate strands of hair. Composition should feel like a dramatic album-cover crop: dynamic diagonal posture, face near the upper right third, ponytail extending left, bib partially cut off at the bottom. Render with clean high-detail Japanese anime painting, luminous eyes, fine linework, soft watercolor-like shading, glossy sweat highlights, and a fresh athletic atmosphere. No extra text, no logo, no watermark.
使用此提示词 ↗
Social Media Post - Anime Girl Cooking Prunes — output

Social Media Post - Anime Girl Cooking Prunes

篝火(かがりび)@AIイラストGPT Image 2
Create a polished vertical anime illustration of {argument name="character name" default="Ena"}, a cheerful young woman cooking prunes in a bright modern kitchen. She has long flowing {argument name="hair color" default="honey-blonde hair"}, large vivid blue eyes, soft blush, and a gentle happy smile. She wears exactly 5 visible outfit pieces: a cream knit cardigan, a light blue vertically striped button-up blouse, a navy knee-length skirt, pink house slippers, and no visible jewelry. Show her standing at a stainless-steel kitchen island with an induction cooktop, holding a wooden spoon above a small silver saucepan filled with glossy dark purple prune compote; include subtle steam and a red heating ring under the pot. In the foreground, place exactly 9 visible loose dried fruits around the bowl and cooktop: 4 dark purple prunes, 3 brown dried prunes or dates, and 2 reddish-brown wrinkled dried fruits. Add one clear glass bowl filled with mixed dried prunes and brown dried fruit on the counter. The kitchen should be clean, white, and sunlit, with tiled walls, white cabinetry, a large window on the left, soft greenery outside, exactly 3 small potted herbs on the windowsill, a sink and faucet beneath the window, a stainless range hood on the upper right, and open shelves holding plates, bowls, mugs, a wooden cutting board, and small jars. Include a small cutting board with a knife on the right counter. Use delicate line art, soft cel shading, pastel colors, realistic kitchen perspective, warm natural morning light, and high-detail anime rendering. Vertical 9:16 composition, waist-to-full-body view, no text, no watermark, no extra characters.
使用此提示词 ↗
Comic / Storyboard - Four-Panel Study Cat Manga — output

Comic / Storyboard - Four-Panel Study Cat Manga

OtntnbGPT Image 2
Goal: Create a clean anime-style vertical 4-panel Japanese manga comic about a schoolboy trying to study while a cat interrupts him, ending with him happily petting the cat instead of studying.

Canvas: Vertical manga page, white outer margin, four horizontal rectangular panels stacked top to bottom with thick black borders and narrow white gutters. Soft cel-shaded anime coloring, simple clean line art, cozy bedroom setting, slightly polished AI-anime look.

Characters: A teenage Japanese schoolboy named {argument name="character name" default="unnamed boy"}, short black hair, navy gakuran school uniform with gold buttons, expressive face. A white domestic cat with pink inner ears and pale yellow eyes, cute and mischievous.

Setting: Same bedroom desk in all 4 panels. Wooden desk at left/center, gray adjustable desk lamp on the left, open spiral notebook and pencil on the desk, bed with pale blue blanket and pillow in the background, light blue curtain, beige wall, blue desk chair. Keep camera mostly side-facing and consistent across panels.

Panel count and layout: Exactly 4 panels.
1. Panel 1: The boy sits at the desk writing seriously in a spiral notebook with a pencil, looking worried and sweating slightly. The desk lamp shines over the desk. Add one thought bubble at the upper right containing {argument name="panel 1 thought" default="明日のテスト、ヤバいな…"}. Add three small writing sound effects near the notebook, reading 「カリ…」, 「カリ…」, 「カリッ…」.
2. Panel 2: The white cat suddenly jumps or climbs onto the open notebook, stretching across the page and blocking the boy's writing. The boy recoils in surprise with mouth open and blush/sweat. Add one jagged speech bubble at the right containing {argument name="panel 2 shout" default="わっ!?"}. Add one rounded speech bubble near the cat containing 「なでて〜」. Add one large cat sound effect near the cat reading 「ぬんっ」.
3. Panel 3: The cat sits upright on the notebook facing the boy. The boy leans forward with both cheeks resting in his hands, looking conflicted and slightly annoyed but softening. Add one thought bubble at the right containing {argument name="panel 3 thought" default="…どいてくれないかな"}. Add one small vertical speech bubble between boy and cat containing 「……」.
4. Panel 4: The boy gives in and pets the cat, smiling warmly with blush while the cat lies curled contentedly on top of the notebook. Pencil lies unused on the desk. Add one speech bubble near the boy containing {argument name="panel 4 line" default="…ま、いっか。"}. Add two petting/satisfaction sound effects: 「ぐ!」 near the cat's head and 「すやすや」 near the sleeping cat.

Visible text count: Include exactly 12 visible text elements total: 4 main boy bubbles/thoughts, 1 cat speech bubble, 1 ellipsis bubble, 3 writing SFX, 1 cat entrance SFX, and 2 final petting/sleeping SFX. Use Japanese vertical manga lettering inside bubbles where appropriate.

Visual style: Cute slice-of-life school manga, clear readable panel composition, expressive faces, modest motion lines, light shadows, warm indoor colors, no halftone texture, no heavy rendering, no watermark, no extra panels, no extra characters.
使用此提示词 ↗
E-commerce Main Image - Realistic Shoyu Ramen Bowl — output

E-commerce Main Image - Realistic Shoyu Ramen Bowl

OGIZARU_おぎざる_GPT Image 2
Create a square, ultra-realistic food photograph of a steaming bowl of {argument name="dish" default="shoyu ramen"} on a rustic wooden table in a cozy ramen shop. The camera is close to the bowl at a slightly elevated front angle, using shallow depth of field: the front ramen toppings and bowl rim are crisp while the rear of the bowl, spoon, and warm wood background fall into soft bokeh. Use natural window light from the upper left, warm brown and golden tones, glossy broth highlights, subtle steam haze, and a high-clarity modern smartphone-photo look. The ramen is served in one dark speckled ceramic bowl with an earthy black-brown glaze and tan rim rings. Include exactly 7 visible topping or serving elements: 1 large sheet of dark green nori standing vertically on the left, 1 halved soft-boiled egg with glossy orange yolk at front left, 1 large round slice of pale chashu pork at the front, 1 small white narutomaki fish cake with a pink spiral near the spoon, 3 tan bamboo shoot strips across the back-left noodles, 1 mound of sliced green scallions in the center, and 1 pale ceramic-and-wood ramen spoon resting across the right rear rim filled with amber broth. Show thin yellow noodles curling through clear brown soy broth with small oil droplets. Background is only the textured wooden tabletop and softly blurred warm interior tones; no people, no text, no logos, no extra utensils, no chopsticks.
使用此提示词 ↗
Comic / Storyboard - Anime Dawn at Kuala Lumpur Sentral — output

Comic / Storyboard - Anime Dawn at Kuala Lumpur Sentral

徘徊家出-はいかいいえで-GPT Image 2
Create a tall vertical anime-style cinematic cityscape at blue hour just before sunrise, showing {argument name="location" default="Kuala Lumpur Sentral"} from an elevated viewpoint. The composition is a 9:16 portrait frame with the lower half filled by a multi-level glass-and-steel transit mall/station complex and elevated walkways, and the upper half dominated by a vast deep indigo sky. Include exactly 3 prominent readable illuminated signs on the station facade: “nu sentral” in white lowercase letters with a blue glow, “KUALA LUMPUR” on a blue sign panel, and “SENTRAL” on a matching blue sign panel. Behind the station, show about 10 dark high-rise towers of varying heights with many tiny warm yellow lit windows. The sky should contain a small thin crescent moon near the upper center-left and hundreds of tiny stars, with dark painterly purple-blue cloud streaks and faint pink dawn light near the horizon. The station interior should glow with warm amber light, revealing silhouettes of commuters and structural beams through large glass panels. Add wet reflective surfaces in the foreground, with orange, blue, and yellow light reflections shimmering across dark water or rain-slick pavement. Use {argument name="visual style" default="detailed Japanese anime background art"}, rich contrast, crisp architectural linework, moody nocturnal colors, glowing windows, subtle atmospheric haze, and a quiet melancholic dawn mood. No characters in close-up, no cars as the main subject, no watermark, no extra text beyond the 3 specified signs.
使用此提示词 ↗
Comic / Storyboard - Anime Girl Holding Tabby Cat — output

Comic / Storyboard - Anime Girl Holding Tabby Cat

Harf_DoneGPT Image 2
Create a polished anime-style vertical illustration of {argument name="character name" default="a young woman"} standing in a cozy apartment bedroom at night, gently lifting and holding {argument name="pet description" default="a gray tabby cat with dark stripes"} in both hands at chest height. She has {argument name="hair color and style" default="long black hair in two loose braids with soft bangs and wispy side strands"}, large expressive brown eyes, delicate features, light blush, and round thin-rimmed glasses. She wears an oversized {argument name="shirt color" default="pale pink"} T-shirt and short gray lounge shorts with white trim, creating a relaxed indoor outfit. Her pose is tender and natural, arms raised as she looks affectionately at the cat; the cat hangs calmly with its tail down, body slightly curved, paws relaxed, and face turned toward her. The setting is a softly lit modern room with a tall dark window on the right showing a nighttime cityscape with blurred building lights, a wooden windowsill with exactly 2 small potted plants, a bookshelf on the left with books and decor, and a larger leafy plant in the background. Use warm indoor lighting from the left/front mixed with cool blue night light from the window, soft shadows, subtle rim light on hair and cat fur, detailed fabric folds, realistic cat fur texture, clean linework, high-quality Japanese anime rendering, romantic quiet slice-of-life mood, portrait composition, waist-up to mid-thigh framing, no visible text except a tiny artist-signature-like mark in the lower right.
使用此提示词 ↗
Profile / Avatar - Warm Backlit Cinematic Portrait — output

Profile / Avatar - Warm Backlit Cinematic Portrait

lizikkGPT Image 2
Create a close-up cinematic portrait photograph of {argument name="subject" default="a young East Asian woman"} turning her head over her bare shoulder and looking directly into the camera with a calm, slightly introspective expression. Frame her from the upper chest to the top of the head in a vertical 4:5 composition, face slightly left of center, shoulder in the foreground, neck elongated, head tilted subtly downward and back toward the viewer. She has {argument name="hair style" default="dark brown-black hair loosely pinned up in a messy bun"} with fine stray strands across the forehead, temples, cheek, and nape, catching the light. Define the facial structure clearly: oval face, soft jawline, delicate cheekbones, straight natural brows, almond-shaped dark brown eyes with small catchlights, a softly rounded nose bridge and tip, natural full lips, visible skin pores, faint under-eye texture, tiny beauty marks on the cheek and neck, and realistic uneven skin sheen. Use {argument name="lighting" default="warm late-afternoon directional backlight from behind the right side, with soft chiaroscuro across the face"}: a golden rim light outlining the hair, cheek, shoulder, and neck, gentle shadow on the far side of the face, specular highlights on the eyes and lips, and no flat frontal lighting. She wears {argument name="clothing" default="a thin gray spaghetti-strap top"}, only the strap and bare shoulder visible. Background is an indistinct warm gray-brown interior or shaded outdoor wall, heavily blurred with shallow depth of field, subtle haze, soft bokeh, and a filmic natural color grade. Make it look like an intimate 85mm lens portrait on full-frame camera, very shallow depth of field, natural grain, imperfect organic realism, fine flyaway hairs, no plastic skin, no overly clean beauty retouching, no text, no watermark, no extra people.
使用此提示词 ↗
Comic / Storyboard - Childlike Koala Hanging From Tree — output
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 ↗