GPT Image 2 prompts

[The Shopkeeper from Another World]
A fire that cannot return, lit in the capital of Slovenia.

{argument name="protagonist" default="Mireia, a dragon-tribe girl from the land of flame scales, Nevia"}.
While escaping from a crumbling volcanic temple, she was sucked into a burning mirror and transported to {argument name="location" default="the riverside of Ljubljana, the capital of Slovenia"}.

Although she hasn't found a way back yet, she has hands that can handle fire.

On the shelves are:
{argument name="product" default="Magic lamps that don't cool down even at night"},
small fire bottles that glow blue when touched,
glass candle holders that mimic dragon scales,
and tinderboxes shaped like old keys.

The old lady, the boy, and the passing painter all smile a little when they see the light.

So today, she opens her shop in this city again.

Use the input image as a {argument name="character" default="character"} reference. Create a photo-style image with a 3:4 aspect ratio of a silver potato chip bag of a fictional brand placed on a realistic wooden table. Convert the character from the input image into a black-and-white hand-drawn manga-style illustration that appears directly on the silver bag. The illustration should occupy a large portion of the empty space on the bottom half of the bag. Maintain the presence of the face, hairstyle, outfit, characteristic decorations, and any partner characters. The potato chips must be from a fictional brand; do not use real company names, logos, or product designs. Determine the product name for the potato chips based on the motif, atmosphere, race, outfit, and partner elements of the character in the input image. The product name should be short, memorable, and sound like something you would see in a convenience store. Use short names like '{argument name="product name" default="OO Potato"}', 'OO Chips', or 'OO Age' whenever possible. The flavor name for the potato chips should also be naturally determined based on the character's impression, choosing from {argument name="flavor" default="Japanese-style, salty, seaweed, black pepper, dashi, or sweet and spicy"} flavors that fit Japanese snacks. The top half of the bag should feature the product package design with a large product name, flavor name, catchphrase, bold black-and-white text, hand-drawn speech bubbles, and playful star or doodle decorations. The bottom half should feature the black-and-white hand-drawn character illustration. The bag should look like a Japanese snack with metallic silver film, wrinkles, reflections, bold printed text, and limited-edition collaboration style decorations. The background and bag should be realistic, while only the character on the bag is in a rough pen, manga line art, or hand-drawn sketch style. The overall atmosphere should be cute, playful, and natural-looking like a collaboration product. Create the image.

Graffiti-inspired sketch animation, expressive ink scribbles, loose gestural drawing, energetic spray paint textures, constantly morphing sketch lines, exaggerated squash and stretch, elastic cartoon physics, playful visual chaos, dynamic hand-drawn motion, lively brush strokes, abstract character animation, spontaneous artistic imperfections, fast-paced slapstick energy, vibrant urban art aesthetic.

[Subject] Refer to Image 1 and Image 2 [Signature] At the bottom right, there is a small Threads logo operated by Meta, and above it, "{argument name="signature" default="CHAse"}" is written in small white cursive like a signature.

[Pose and Expression]
A portrait pose looking directly at the camera with the chin resting on the palm of the right hand. The intense and deep gaze through the stray hairs overwhelms the viewer, and the serious and dreamy expression with slightly closed lips exudes a romantic yet lonely charisma.

[Clothing]
A {argument name="hat style" default="wide-brim cowboy hat"} with unique zigzag stitch decorations along the edge is pressed down low on the head. The subject wears a white blouse, showing a natural Western-inspired look.

[Hairstyle and Makeup Details]
Long, wavy hair with a messy, roughly groomed texture. Strands of stray hair naturally cover the face line and eyes, creating a raw and free mood. The makeup is a minimal look that preserves freckles and the skin's natural texture, with the depth of the lips and eyes emphasized in the black-and-white contrast.

[Lighting and Direction]
Dramatic backlighting coming strongly from the back-left of the figure is the main light source. This intense light creates a dazzling rim light highlight along the outline of the messy hair and the edge of the hat brim, while the front of the face is submerged in soft shadow, maximizing the depth and three-dimensionality of the features.

[Texture and Color Mood]
Expressed in a classic {argument name="color tone" default="Monochrome"}, focusing entirely on the contrast of light and shadow without color interference. The fine texture of thick hair, the matte and soft leather or felt texture of the hat, and the smooth skin texture contrast vividly within a deep grayscale, creating an artistic and sensual fine-art pictorial mood.

[Film and Camera Specs]
Shot using a f/2.0 aperture, 1/250s shutter speed, ISO 400, and an 85mm telephoto prime lens. A tight close-up shot captured at eye level to fully immerse in the subject's intense gaze and expression. Shallow depth of field is applied, focusing sharply on the eyes and front of the face while softly blurring the back of the hat and the ends of the hair. The subtle film grain characteristic of high-sensitivity black-and-white film completes the classic cinematic sensibility.

[Background (Optional)]
A minimal, flat, light-toned background with no details to highlight the silhouette of the subject and hat. By excluding complex visual elements, the focus is entirely on the emotional line and formal beauty of the foreground figure.

Set aspect ratio to 1:1.

A cinematic close-up portrait of a rugged, sophisticated middle-aged man with thick wavy dark brown hair streaked with subtle gray, a full well-groomed salt-and-pepper beard, wearing premium {argument name="eyewear style" default="translucent amber acetate round sunglasses"} with {argument name="lens color" default="dark green"} lenses, dressed in a {argument name="outfit" default="tailored navy blazer over a crisp white open-collar shirt"}. Soft natural window light from the side creates gentle contrast, highlighting skin texture and facial details. Moody luxury editorial aesthetic, timeless European style, shallow depth of field, blurred urban background, rich earthy color grading, ultra-realistic skin, premium fashion campaign, photorealistic, 85mm lens, f/1.8, high dynamic range, razor-sharp focus on the face and sunglasses, subtle film grain, luxury eyewear advertisement, 8K, highly detailed.

Mixed media collage animation style, a stylish young woman walking confidently toward the camera, realistic photographic face and skin, shoulder-length softly wavy {argument name="hair color" default="auburn"} hair, defined natural brows, almond-shaped eyes, refined feminine features, delicate gold hoop earrings, wearing a clean white {argument name="clothing" default="T-shirt"}, outlined with glowing 2D neon green and hot pink scribble lines, {argument name="location" default="New York"} background with recognizable urban city energy, skyscrapers, street elements, fire escapes, layered city depth, floating 3D chrome liquid shapes in the air, grungy edges, acid graphics aesthetic, dynamic fisheye lens perspective, vibrant juicy colors, high contrast, ultra-detailed, trendy editorial collage feel, stop-motion feel, bold modern composition --ar 4:5 --raw --profile glu65pn --stylize 60 --weird 4 --hd

Create a stylized collectible toy package featuring {argument name="character name" default="[CHARACTER NAME]"}.

Packaging color: {argument name="packaging color" default="[COLOR]"}
Top text: "{argument name="title" default="[TITLE]"}"
Figure style: adorable vinyl designer toy, oversized head, simplified proportions.
Pose: [POSE]
Expression: [EXPRESSION]

Accessories:
• [ITEM 1]
• [ITEM 2]
• [ITEM 3]
• [ITEM 4]
• [ITEM 5]

Outfit:
[OUTFIT DESCRIPTION]

Professional product photography, molded plastic tray, luxury collectible packaging, front view, symmetrical layout, soft shadows, matte finish, ultra-detailed 3D render.

A vibrant, highly detailed, ultra-realistic photorealistic image of a playful {argument name="scene" default="springtime scene"}. A beautiful young woman stands outside a house with smooth beige walls, making a 'shh' secret gesture with her finger to her lips while playfully hiding a {argument name="object" default="bouquet of pink flowers"} behind her back. She has wavy black hair and is wearing a light pink long-sleeved blouse with a blue bow tie, a high-waisted pleated blue midi skirt with thin white stripes, and blue canvas sneakers. Beside her, a {argument name="animal" default="fluffy orange tabby cat"} sits patiently on the exterior ledge of a blue-framed window. Inside the window, a smiling young man with short Black hair, a slight beard, and a teal jacket looks out happily. The scene is bathed in bright, sunny daylight, casting distinct, beautiful silhouettes of tree branches onto the wall of the house. Soft, out-of-focus pink blossoming branches frame the bottom foreground, adding a warm depth of field. Clean, bright, romantic, cinematic lighting, hyper-realistic.

Create a candid flash photograph of two stylish friends posing closely at an indoor party or lounge, shot from slightly above with a wide-angle lens and warm amber color grading. The scene shows exactly 2 people: on the left, {argument name="woman outfit" default="a woman in a cream one-shoulder fitted dress with a sculptural rosette at the chest, long straight black hair, small earrings, and dark oval sunglasses"}, leaning forward with one hand resting against her cheek; on the right, {argument name="man outfit" default="a man in a black short-sleeve polo shirt with curly reddish-brown hair and small dark sunglasses"}, leaning in with one hand behind his back. Cover each person's face with a centered opaque square privacy block in a warm tan-brown color, exactly 2 face blocks total, leaving hair, sunglasses edges, shoulders, and clothing visible. The background is a cozy retro interior with vertical wood paneling, beige curtains, a red curtain at the left, reddish carpet, and large glowing white spherical pendant lights near the top edge, including exactly 4 visible globe lights: one cropped at the top-left corner, one large central globe, and two overlapping globes at the top-right. Use direct on-camera flash, high ISO party-photo texture, slight motion softness, realistic skin tones, strong warm highlights, intimate celebrity-event snapshot feel, square crop, no text, no watermark.

Goal: Create a whimsical cinematic luxury ice-blue café scene starring {argument name="character name" default="Ingrid"}, using the uploaded person as the exact face reference if provided, with face identity locked and repeated consistently across every character version.

Canvas: Square 4:5-feeling vertical composition, ultra-detailed glossy 3D illustration / cinematic digital art, icy blue and white palette, sparkling frost, snowflakes, cool mist, reflective frozen marble floor, bright rim lighting, and a magical winter café atmosphere.

Main subject: Center the main {argument name="character name" default="Ingrid"} seated inside an oversized glossy white ceramic mug chair shaped like a giant coffee cup. The mug has a large handle on the right and is labeled in elegant black serif text: "INGRID" with smaller text below: "• COFFEE QUEEN •". She wears a cozy oversized {argument name="cardigan color" default="pale ice-blue"} knitted cardigan, white top, blue jeans, and holds a hot coffee cup with both hands. Her long wavy brown hair flows naturally, and she sits on soft pale-blue cushions inside the cup.

Environment and signage: Set the scene inside a premium frozen café called "INGRID'S ICE STORAGE — CAFÉ —" in large glowing white-blue letters on the back wall. Add icicles hanging from the ceiling edges, frosted glass walls, glowing display cases, and a circular ceiling light above the main subject casting radiant blue beams downward. Include these visible signs exactly: "PREMIUM ICE CREAM" above the left freezer shelves, "FROZEN FRUITS" on the lower-left case, "COFFEE & MORE" above the right coffee station, "INGRID'S CAFÉ" on the black coffee menu board, "ICED DRINKS" on the lower-right drink case, and "DESSERTS & TREATS" on the dessert display.

Repeated character count: Show exactly 7 total versions of {argument name="character name" default="Ingrid"}: 1 large central Ingrid in the giant mug chair; 1 mini Ingrid sitting on the left ice-cream shelf holding a drink; 1 mini Ingrid standing at the right coffee machine; 1 mini Ingrid sitting lower left reading a menu and drinking coffee; 1 mini Ingrid sitting near the bottom-left foreground holding a soft-serve ice cream cone and wearing an Ingrid café apron; 1 mini Ingrid sitting bottom center with a small shopping basket of fruit; and 1 mini Ingrid sitting at a small round café table on the right holding coffee.

Penguin count: Include exactly 5 cute plush-like penguins wearing blue winter hats and scarves: 1 on the left shelf beside the mini Ingrid; 1 small penguin near the lower center-left beside the reader; 1 large penguin in the bottom-left foreground; 1 small penguin near the right side of the giant mug; and 1 penguin beside the right café table.

Food and drink details: Add frozen shelves filled with branded ice-cream tubs labeled with the character name, iced drinks in bottles and cups, desserts in glass cases, whipped cream treats, pastries, and colorful frozen fruit. Include a fruit basket with pineapple, oranges, lemons, and berries near the bottom center. The right black coffee menu board should show exactly 6 illustrated drink options labeled: Americano, Latte, Cappuccino, Mocha, Vanilla Latte, and Caramel Macchiato.

Style: Hyper-detailed, glossy, high-end fantasy café advertising look, rich reflections, soft volumetric light, blue-white neon glow, snow particles, frosty glass, polished ceramic, cozy knit textures, cute miniature characters, and charming winter mascot energy.

Constraints: Keep all faces consistent with the reference person if one is provided. Do not add extra character copies beyond the 7 specified, and do not add extra penguins beyond the 5 specified. Keep the text legible and centered where described. No watermark, no logo, no distorted hands, no unreadable main signage.

Goal: Create a vertical 3×3 lifestyle photo collage for social media, combining candid streetwear OOTD, city-walk travel details, and playful white hand-drawn diary annotations.

Canvas: Vertical 9:16 smartphone story format, about 504×896 px. Use a casual urban travel aesthetic with muted colors, natural daylight, slight film grain, and realistic smartphone photography.

Layout: Arrange exactly 9 photo panels in a tight grid with thin or no gutters. The panels should feel like a curated travel朋友圈 / Instagram story collage.

Subject details: Feature one young adult man, {argument name="main subject" default="a slim young man in a navy baseball cap, dark navy overshirt jacket, white T-shirt, loose khaki trousers, black canvas sneakers with white laces, and a black crossbody bag"}. He appears in multiple candid poses: walking, standing with a drink, looking at his phone, and shown from his own POV.

Panel count and contents: Include exactly 9 panels:
1. Top-left: full-body side view of the man walking beside a brick wall and dark shopfront, slightly hunched with hands near pockets; add white doodles reading “city walk” at top left and “day off” near bottom left, with an outlined silhouette around his body, an arrow near his shoes, and a small star.
2. Top-center: close-up of blue bilingual street direction signs on a pole in a city street, with brick buildings and trees behind; add curved white arrows, a small heart, emphasis marks, and the handwritten words “getting lost” near the lower right.
3. Top-right: medium shot of the man standing outside a brick storefront holding an iced coffee; add a white outline around his silhouette, a curved arrow pointing to the drink, handwritten text “coffee time,” and a small star.
4. Middle-left: close-up of the crossbody strap and denim/navy fabric with a small plush keychain charm; add sparkles, a curved arrow, and handwritten text “little guy :)”.
5. Middle-center: downward POV of khaki pants and black sneakers on grey pavement while walking; add white oval motion doodles around both shoes and text “step by step”.
6. Middle-right: street corner view with a red-brick building, trees, crosswalk, and blue sky; add a thick white cloud doodle along the bottom edge, a white wavy line tracing the rooftop, a small heart, and handwritten text “nice view” with an arrow.
7. Bottom-left: candid full-body/three-quarter shot of the man standing on a sidewalk looking down at his phone, with street signs and a red post box or utility box in the background; add sweeping white line doodles and text “on my way”.
8. Bottom-center: close-up of a white takeaway cup with red straw and large red letters, held while walking; add a white circle around the cup, small accent lines, a cloud-shaped white label at the bottom reading “cheers!”, and a tiny heart.
9. Bottom-right: downward seated or standing POV of sneakers, khaki pants, and dark jacket on asphalt with painted road markings; add a white outline around the shoes, a large hand-drawn road-line shape, a small star, and text “keep going” with an arrow.

Visual style: Realistic candid photography mixed with hand-drawn white marker graphics. Use high-contrast white doodles, casual handwritten lettering, arrows, hearts, sparkles, cloud cutouts, curved lines, and silhouette outlines. Overall mood should be relaxed, aesthetic, urban, and travel-diary-like.

Setting: Use {argument name="city setting" default="a leafy historic brick-building neighborhood with storefronts, road signs, crosswalks, and grey pavement"}. Keep lighting natural and slightly overcast with soft shadows.

Text content: Include these exact 9 handwritten English phrases: “city walk”, “day off”, “getting lost”, “coffee time”, “little guy :)”, “step by step”, “nice view”, “on my way”, “cheers!”, and “keep going”. The street signs may contain small bilingual Chinese/English-style text but should remain secondary and realistic.

Color and mood: Muted navy, khaki, brick red, grey asphalt, soft green trees, white doodles. Make it feel like {argument name="social media mood" default="an effortless weekend city walk travel diary"}.

Constraints: Do not create a polished studio shoot. Avoid extra panels, extra people, heavy filters, neon colors, or excessive text. Keep all doodles white and hand-drawn, integrated naturally over the photos.

Goal: Create a luxurious vertical poster announcing a collection of useful GPT Image 2 prompt websites, styled like a premium treasure discovery.

Canvas: Portrait 3:4 poster, bright white and light silver background, high-resolution commercial 3D render mixed with bold graphic typography. Use a clean editorial layout with subtle grid marks, thin crop lines, dotted ornaments, small sparkle icons, and bottom micro-print details.

Top typography: Place a huge, heavy, blocky black Chinese headline centered near the top reading {argument name="main headline" default="提示词宝藏站"}. Above and around it add minimal black decorative marks: one large four-point star at upper left, thin horizontal lines, a dotted line near upper right, and tiny diamond markers.

Subtitle bar: Directly under the headline, add a thin outlined rectangular label containing bold black text reading {argument name="subtitle text" default="10 个超实用 GPT Image 2 提示词网站"}. Keep the subtitle centered with generous letter spacing and a crisp print-design feel.

Main subject: In the lower two-thirds, show an open futuristic silver treasure chest with polished metal panels, transparent glass-like edges, chrome trim, gold accents, and a large faceted diamond emblem on the front lock. The chest is overflowing with glowing prompt-library interface cards, gems, diamonds, and warm golden light beams. Add white, gold, and crystal reflections, lens flares, and sparkling particles.

Website cards: Show exactly 10 stacked floating UI cards emerging from the chest, arranged in layered perspective like browser panels. Each card should have rounded corners, small app icons, minimal menu dots, and one visible label. The 10 visible labels are: 1) Midjourney Prompt Hub, 2) Prompthero, 3) OpenGPT, 4) LibLib AI, 5) PromptJourney, 6) FlowGPT, 7) PromptBase, 8) Libai.GPT, 9) ChatGPT Prompt Library, 10) Magic Prompts. Use a mix of white, silver, and black card headers.

Search element: In front of the cards, place a pill-shaped search bar angled slightly upward, glowing softly, with the text {argument name="search query" default="GPT Image 2"} and a black search icon button on the right.

Decorative objects: Include exactly 2 translucent ribbon bands looping diagonally through the composition, both printed repeatedly with the word {argument name="ribbon word" default="PROMPTS"}. Add exactly 1 gold ring on the left, exactly 2 large faceted diamonds near the left and right sides, exactly 1 small floating rounded-square tile with a white geometric cube logo on the left, and exactly 1 matching rounded-square tile with a black cube logo on the right. Scatter smaller crystal shards and gold star particles around the chest.

Visual style: Premium tech treasure aesthetic, glossy 3D rendering, white/silver/gold palette, cinematic reflections, realistic glass and metal materials, sharp typography, high contrast black headline, elegant minimal infographic poster styling.

Constraints: Keep the design vertically centered and symmetrical, preserve the exact count of 10 website cards, avoid adding extra website names, avoid people or characters, no watermark, no social media UI, no QR code.

Goal: Create a dark-mode mobile chat interface screenshot showing an AI image-generation result that failed to make four separate images and instead produced one 2x2 collage.

Canvas: Square screenshot, approximately 768×768 px, black background with a cropped mobile-app UI feel. Leave large empty black space on the right side and below the result.

Layout: At the top center, place a dark gray rounded prompt bubble containing the Chinese command text “生成4张独立的图片” with a thick red underline beneath it, and a smaller line below reading “收起 ^”. Below and left, show a status label in Chinese reading “已停止思考 ›”. Under that, place one generated image card with rounded top corners, containing exactly 4 photo panels arranged in a 2×2 grid, separated by thin white grid lines. Add a small circular download button with a downward arrow overlapping the lower-right corner of the collage. At the bottom, show a dark rounded input bar spanning the width: a plus icon on the left, placeholder text “有问题，尽管问”, and on the right the mode text “极速⌄”, a microphone icon, and a white circular voice/equalizer button.

Generated image content: The single generated image card must contain exactly 4 panels: top-left, top-right, bottom-left, bottom-right. Each panel shows the same young woman posing in front of a brightly lit juice shop or drink stand at night, with warm storefront lighting, neon signs, colorful drink menu boards, and a shallow depth-of-field smartphone-photo look. She wears a pink sleeveless crop top and a white pleated mini skirt, with hair styled in two cute buns. Her face is obscured by a soft rectangular blur for privacy. In each panel she holds a different plastic drink cup with a straw, posing slightly differently: 1) top-left, holding a red drink with both hands near her chest; 2) top-right, leaning forward holding a yellow/orange drink near her face beside neon “JUICE” signage; 3) bottom-left, holding a large orange drink close to her mouth in front of a bright menu board; 4) bottom-right, holding a pale pink drink while standing near a glass display case. The subject should be {argument name="character description" default="a young woman with two hair buns, pink sleeveless crop top, white pleated mini skirt, face blurred"}, photographed at {argument name="location" default="a colorful night-time juice shop storefront"}.

Visual style: Make the overall screenshot look like a real captured phone UI, not a polished mockup. Use slightly compressed screenshot quality, soft glow from neon lights, realistic photo colors in the collage, and minimalist dark UI elements. The Chinese command in the top bubble should be {argument name="command text" default="生成4张独立的图片"}. The status label should be {argument name="status text" default="已停止思考 ›"}. The bottom input placeholder should be {argument name="input placeholder" default="有问题，尽管问"}.

Constraints: The result must show one combined 2×2 collage rather than four separate image cards. Include exactly 4 photo panels and exactly 1 download icon. Preserve the dark chat interface, Chinese UI text, red underline, and bottom input bar. Do not add extra panels, extra people, watermarks, or captions inside the generated photo collage.

Create a wide sticker-style fan poster featuring {argument name="character name" default="a young male streetwear model"} sitting casually in the center on a clean white background. The central figure is a black-and-white cutout photo effect with a deliberately blurred/obscured face, wearing a black baseball cap, black oversized T-shirt, loose dark jeans, white socks, black-and-white sneakers, a chain necklace, and a wristwatch; pose him seated on the floor with legs spread forward, one elbow resting on a knee and one hand near his head, relaxed idol streetwear energy. Surround him with bold blue, orange, pink, and black hand-drawn doodle stickers and graffiti typography in a playful Gen-Z aesthetic. Include exactly 17 visible doodle/text elements around the person: 1 large blue graffiti word “OOTD” on the left with a black shadow and a small black ribbon underneath reading “OUTFIT OF THE DAY”; 2 a blue crown above the OOTD; 3 a small “GOOD VIBES” note with a pink heart near the top center; 4 an orange comic burst on the upper right containing “COOL”; 5 a large blue oval sticker on the right reading “TODAY” with a pink heart; 6 black handwritten “STYLE” with pink “UP!” beneath it on the lower right; 7 a large blue star at bottom right; 8 a blue lightning bolt at top right; 9 an orange spiral squiggle at the far upper right; 10 a small black squiggle near the right edge; 11 a black curved arrow pointing upward near the right middle; 12 a loose blue wavy scribble near the lower right center; 13 a blue underline scribble beneath the shoes; 14 an orange speech bubble on the lower left reading “NICE!”; 15 black text “BE YOURSELF” with a blue underline on the left; 16 a blue outlined star on the left; 17 a pink lightning doodle near the lower left. Add a few short black motion marks around the stickers but do not add extra labeled stickers. Use thick marker outlines, energetic hand lettering, high contrast, crisp cutout edges, slight gray floor shadow under the seated person, clean modern collage composition, blue-green/blue-dominant sticker vibe with orange and pink accents. Keep the canvas 16:9 landscape, poster-ready, no watermark, no border.

Create a vertical 2:3 high-detail anime fantasy illustration of {argument name="character name" default="a silver-haired battle maiden"} in a ruined gothic cathedral during combat. The character is centered in a dynamic low-angle three-quarter pose, leaning forward with one arm extended to the right holding a huge translucent circular shield or magical glass barrier, and the other arm lowered to the left with a heavy mechanical gauntlet cannon emitting smoke and orange heat. She has long flowing {argument name="hair color" default="silver-white"} hair, pale skin, a calm serious expression with half-lidded eyes, and delicate white floral head ornaments with ribbon-like streamers whipping through the air. Her outfit is a dark, ornate, tattered gothic dress with lace trim, layered white underskirt, black bodice, armored sleeves, and frayed cloth strips swirling around her. The environment is a vast collapsed stone cathedral with pointed arches, broken stained-glass windows, crumbling walls, distant ruined towers, dust, fog, rubble, and shafts of cold daylight. The shield is an oversized oval disk of clear glass-like energy with etched concentric circles and radial patterns, catching bright highlights on the right side of the image. Use a monochrome gray, ash, ivory, and charcoal palette with small warm ember sparks and glowing orange accents. Style should be painterly Japanese anime concept art, extremely detailed linework, dramatic motion, cinematic lighting, high contrast, watercolor-and-ink texture, atmospheric smoke, flying debris, no visible text, no watermark, no border.

Goal: Create a playful summer lifestyle photo collage featuring {argument name="character description" default="a young woman with short dark hair wearing a blue-and-white striped sleeveless top and light shorts"} enjoying watermelon in a sunny park, decorated with hand-drawn white doodles and casual handwritten English words.

Canvas: Wide horizontal 16:9 image, arranged as an exact 3 by 3 grid of 9 equal rectangular photo panels with thin white borders between panels. Use soft natural daylight, fresh green grass, shallow depth of field, and a bright airy summer color palette.

Layout and panel count: Include exactly 9 panels:
1. Top-left: the woman sits cross-legged on grass holding a large watermelon slice in front of her face, with a white fence and cloudy sky in the background; doodles include a smiley face, short rays, the word “Hi!”, and small splash marks.
2. Top-center: close-up still life of several watermelon slices on a plate or tray beside an orange bag; doodles include a crown-like squiggle, sparkle marks, and the handwritten word “Sweet!”.
3. Top-right: side view of the woman sitting on grass, smiling toward or holding a watermelon slice near her mouth; doodles include “Hello!”, quotation marks, a swirl line, and a heart.
4. Middle-left: close-up portrait crop of the woman holding a small fork or skewer with a tiny watermelon cube; doodles include radiating marks and a large question mark.
5. Middle-center: close-up of the woman holding a watermelon slice near her mouth, face mostly centered and partially obscured by the fruit; doodles include dotted lines, a heart, and a small smiley face.
6. Middle-right: the woman sits on grass holding a watermelon slice up to her face; orange bag nearby; doodles include “Nice Day!”, a wavy underline, quotation marks, short grass strokes, and small sparkle lines.
7. Bottom-left: wider overhead-leaning shot of the woman reclining or sitting on a picnic cloth on grass with watermelon and orange bag; doodles include a curved arrow, smiley face, small hearts, and motion marks.
8. Bottom-center: tight portrait crop of the woman holding an orange hat or orange bag close to her chest or face, with soft grass background; doodles include small hearts, a curved white stroke, and a large number-like squiggle.
9. Bottom-right: overhead shot of the woman lying or reclining on grass with one arm raised, orange bag beside her, and a watermelon slice near her hand; doodles include a speech bubble saying “Yummy!”, small dash marks, and a heart outline.

Subject details: The subject should appear cheerful, relaxed, and candid, like a summer picnic travel-photo shoot. Keep the face softly anonymized or blurred, while preserving the pose, hair silhouette, outfit, and body language. Use {argument name="main prop" default="fresh red watermelon slices with black seeds"} as the recurring prop in multiple panels, and include {argument name="accent accessory" default="a bright orange tote bag or sun hat"} in several panels.

Visual style: Realistic photography, soft film-like tones, slightly overexposed summer highlights, green park setting, casual social-media collage aesthetic. Add white hand-drawn doodles as if sketched with a marker on top of the photos: hearts, arrows, smiley faces, sparkles, squiggles, quotation marks, and short handwritten words.

Text content: Use exactly these visible handwritten English words across the collage: {argument name="greeting text" default="Hi!"}, “Sweet!”, “Hello!”, “Nice Day!”, and “Yummy!”. Keep the lettering informal, white, and doodle-like.

Constraints: Maintain the exact 9-panel grid, thin white dividers, no extra panels, no watermark, no app interface, no captions outside the collage, and avoid adding additional readable words beyond the specified doodle text.

Create a fisheye, inside-the-washing-machine point-of-view fashion photo of {argument name="character name" default="a young woman"} crouching directly in front of an open laundromat washer, reaching both hands toward the circular metal rim as if looking into the drum. The image should be framed by a thick dark circular washer-door border with chrome highlights, strong wide-angle distortion, warm late-afternoon sunlight, beige tile floor, and a vintage coin-laundry background. Show exactly 3 front-loading washing machines across the back wall: one large chrome round door on the left, one mostly hidden behind the subject in the center, and one large chrome round door on the right. Above them, show exactly 3 readable wall signs: a left sign labeled “LAUNDRY INSTRUCTIONS,” a large center sign reading “TRY OUR GIANT WASHER FOR COMFORTERS TOO!,” and a right sign listing “SOAP BLEACH SOFTENERS.” The subject has {argument name="hair color" default="long wavy honey-brown hair"}, a casual Y2K streetwear look, and a deliberately obscured face covered by a centered flat tan rectangular censor block. Outfit details: exactly 5 visible clothing/accessory pieces — a cream ribbed cropped tank top with red trim and the word {argument name="tank top text" default="Honey"} plus two small cherry graphics, loose red pants with white star graphics down the side, red-and-white sneakers, a small star pendant necklace, and small earrings. Add subtle film grain, golden nostalgic color grading, realistic reflections in the washer doors, slight lens softness at the edges, and faint semi-transparent watermark text around the vertical edges and across the frame reading {argument name="watermark text" default="♥ NaEun & Damien 2026"}. Keep the composition intimate, low-angle, centered, photorealistic, and editorial, with the subject’s arms foreshortened toward the viewer and the laundromat signage clearly visible.

Create a vertical 4:5 high-fashion editorial magazine cover portrait with a quiet old-newspaper mood. A young woman with {argument name="hair color" default="long dark wavy hair"} sits on the floor against a plain warm gray studio wall, positioned slightly right of center with lots of negative space on the left for typography. She has a calm, unspoken, introspective gaze and a soft natural expression. She wears loose vintage blue jeans and is otherwise bare-shouldered, modestly covering her chest by holding a folded gray newspaper across her torso with one arm wrapped around it; the newspaper has dense English column text. Her pose is relaxed and asymmetrical: one leg bent forward across the foreground, the other folded underneath, one hand braced on the floor to her right, bare feet visible. Use muted gray tones, soft diffused side lighting, gentle shadows on the floor, subtle film grain, desaturated colors, realistic skin texture, and an elegant minimalist fashion-photography finish.

Magazine cover typography: use exactly 5 text blocks in refined black serif type, arranged like an art magazine/newspaper cover. Block 1: large uppercase headline at the upper left reading {argument name="headline text" default="WHISPERS\nUNSPOKEN"}. Block 2: small italic subheading beneath it, preceded by a thin vertical rule, reading {argument name="subheading text" default="Some feelings never\nmade it to words."}. Block 3: small issue/date text below, reading {argument name="issue text" default="ISSUE No.27\nMAY 2025"}, with a short horizontal line underneath. Block 4: mid-left stacked title reading {argument name="side title" default="SILENCE,\nIN\nBETWEEN\nUS."}, with a tiny italic caption below reading “A moment held between truth and hesitation.” and a short horizontal rule beneath. Block 5: tiny footer at bottom left reading {argument name="footer text" default="PHOTOGRAPHED\nIN SEOUL"}. Keep the layout airy and sophisticated, avoid bright colors, avoid extra text, avoid logos or watermarks, and make the final image look like a premium quiet fashion magazine cover printed on aged gray paper.

Create a photorealistic vertical street-fashion snapshot in Tokyo, Japan, showing {argument name="character name" default="an adult woman"} posing beside a wall of capsule toy vending machines on a busy pedestrian shopping street. She is the single main subject, leaning forward slightly toward the camera while turning a gashapon knob with her right hand and touching her hair near her ear with her left hand. Her face is intentionally hidden by a large soft-edged square blur/censor block in muted brown tones. She has long dark brown hair and wears a light blue button-up blouse with rolled sleeves, a dark navy bow tied at the chest, a very short navy pleated skirt, black knee-high socks, and white sneakers. Add exactly 3 visible accessories on her: a large navy shoulder bag, a small white teddy bear keychain hanging from the bag, and two wrist scrunchies on the left wrist, one pink flower-like scrunchie and one black scrunchie. The right side of the image contains exactly 6 prominent white capsule toy machine fronts in a vertical perspective row, with colorful anime-style product posters, coin slots, clear capsule windows, and blue-and-white turning knobs; her right fingertips rest on one knob. The background is a crowded Akihabara/Shibuya-style sidewalk with many pedestrians walking away, bright Japanese storefront signs, colorful billboards, and dense urban advertising; keep the background slightly less sharp than the subject. Use bright natural daylight, realistic skin and fabric texture, smartphone camera perspective, wide-angle 3/4 full-body framing from slightly above, vibrant colors, high detail, candid social-media photo realism. Customize the outfit color as {argument name="outfit color" default="light blue blouse and navy skirt"}, the location as {argument name="city street" default="Tokyo anime shopping street"}, the machine theme as {argument name="capsule toy theme" default="cute anime character gashapon"}, and the face concealment as {argument name="face effect" default="large square brown blur over the face"}. Avoid extra limbs, distorted hands, unreadable anatomy, duplicate bags, or additional main subjects.

Create a cheerful vertical anime-style illustration of a cute pudding mascot greeting the viewer in a sunny riverside forest. The main subject is a plump custard pudding creature with a glossy caramel top, small rounded limbs, shiny black oval eyes, and a wide happy open mouth; it sits on bright green grass beside a sparkling blue stream. On top of the pudding place exactly 2 dessert toppings: one swirl of white whipped cream and one shiny red cherry with a stem. Add large playful Japanese greeting text across the upper sky reading {argument name="greeting text" default="こんにちは"}, in thick rounded orange lettering with a white outline, decorated with exactly 2 small flower icons: one pink flower at the upper left of the text and one yellow-orange flower at the upper right. Surround the mascot with exactly 4 colorful music-note icons: one orange note near the left upper side, one yellow note at the left middle, one pink note at the right middle, and one yellow note at the upper right. Include small white sparkles and curved magical motion lines around the character. The environment should show a clear blue sky with soft clouds, dense sunlit green trees, a flowing river on the left with foam and glittering highlights, rounded river stones in the foreground, and small yellow and white wildflowers in the grass. Use warm golden sunlight, soft painterly anime rendering, saturated colors, cute kawaii mood, shallow depth of field, polished digital art, no watermark, no extra text beyond the greeting.

Using REFERENCE_0 as the character base, redraw the rough black-and-white sketch into a polished, original anime character design sheet while keeping the same bust-up composition, short bob hairstyle, blazer-and-blouse outfit, and face-obscuring rectangle placement. Convert the sketch into clean expressive ink lines with soft watercolor-style coloring: warm brown hair, natural skin tones, a light beige blazer, white blouse, and a small delicate necklace. Keep the background plain white and expand the canvas horizontally so there is empty space on the left for handwritten character notes.

Add exactly 4 handwritten Japanese text lines on the left side, in black ink with subtle pink underline accents:
1. {argument name="character name" default="桜井 梓（35）"}
2. {argument name="profile line one" default="仕事はデキるけど、"}
3. {argument name="profile line two" default="恋愛はちょっぴり不器用。"}
4. {argument name="profile line three" default="本当は素直になりたい。"}

Style: refined but still sketch-like anime illustration, clean corrected anatomy, loose hand-drawn linework, soft marker/watercolor shading, gentle contemporary character-concept art feel. Do not add a detailed background, props, or extra characters.

Create a square gothic monochrome anime illustration with an antique book-cover feel. The composition shows an androgynous young man with {argument name="hair color" default="pale silver-white"} tousled medium-length hair, wearing a high-collared ruffled blouse, dark ribbon tie, vest-like formal layers, fitted gloves, and buttoned cuffs. He is positioned slightly right of center, leaning forward with one gloved hand raised near his mouth in a secretive, melancholy pose; keep the central face area intentionally obscured by a flat opaque square censor block in warm gray-beige, covering most facial features from forehead to mouth. Behind him on the left is an ornate oval Victorian mirror with a carved baroque frame, reflecting a soft ghostly side profile of the same character. Surround the figure with dense flowers and foliage: large pale orchids in the upper right, many small clustered blossoms across the lower left and center, leaves and buds throughout. Add vertical Japanese title text on the left reading {argument name="vertical title text" default="擬花"}, set in elegant serif/calligraphic characters with thin decorative horizontal flourishes. Use a muted sepia-gray palette, watercolor and ink wash texture, soft shading, high detail, delicate linework, dreamy dark background, and subtle paper grain. Frame the entire image with a thin antique border and ornate corner filigree decorations on all four corners, plus small centered ornamental flourishes at the top and bottom. Canvas: 1:1 square. Constraints: exactly one main character, exactly one mirror reflection, exactly one opaque square over the face, exactly two visible Japanese title characters, and four ornate corner decorations; no modern objects, no bright colors, no watermark.

Create a vertical 9:16 monochrome sepia anime illustration of an androgynous young noble with {argument name="hair color" default="silvery white"} tousled layered hair, posed in a melancholic Victorian-gothic garden scene. The character wears a glossy white ruffled blouse with a high frilled collar, a black ribbon bow at the neck, one visible white glove touching the cheek, and one visible cuff with exactly 3 round buttons. Their long hair is tied into one low side ponytail over the right shoulder. Place one ornate antique hand mirror along the left side, with carved metallic floral scrollwork, reflecting the character’s downcast face and closed or half-lidded eye. Add an opaque centered square censor block in muted warm gray-taupe covering the main face area, about one third of the image width, with sharp clean edges. Surround the portrait with white flowers: exactly 3 large orchid blossoms in the upper background, plus multiple clusters of small five-petal white flowers near the right side and lower left foreground. Use delicate manga linework, painterly grayscale shading, soft highlights, intricate hair strands, romantic dark foliage, shallow depth of field, high detail, elegant sorrowful mood, no text, no watermark.

Create a cinematic two-panel automotive action photograph in a 16:9 horizontal frame, split exactly down the center by a thin vertical black divider. Left panel: an interior cockpit shot from the passenger side of a classic compact sports car, showing one driver wearing a black racing jacket with two white sleeve stripes, black leather driving gloves, a black harness, and a small gold hoop earring. The driver has messy dark hair, their face intentionally obscured by a plain dark brown square, hands gripping the steering wheel, seated in a black RECARO-style bucket seat. The cabin is dark, tight, and realistic, with blurred forest and guardrail streaking past the window to imply speed. Right panel: a dynamic rear three-quarter tracking shot of a {argument name="car model" default="white Toyota MR2 AW11 late-model G-Limited Supercharger"} driving fast on a winding Japanese mountain road. Show the car from low behind, centered slightly right, with rectangular red taillights, black rear trim, rear spoiler, rear glass engine cover area, visible badge text including “TOYOTA MR2” and “SUPER CHARGER,” and a Japanese license plate reading {argument name="license plate text" default="20-00"}. The road curves through dense green forest with metal guardrails, stone retaining walls, orange center lines, and strong motion blur on the asphalt. Use photorealistic DSLR quality, natural overcast daylight, muted green and gray tones, high shutter-speed realism mixed with panning blur, dramatic touge racing atmosphere, and no extra captions, logos, watermarks, or added vehicles.