GPT Image 2 prompts

Create a detailed infographic showing the working principle and process of an automatic coffee machine like {argument name="brand" default="Jura"}.
Cover components from the bean hopper, grinder, scale, water tank, to the boiler.
I want to understand this process both technically and visually.

Subject: {argument name="subject" default="besties portrait"}

Character/Expression: Two young women with translucent skin, bright eyes, natural Japanese-style makeup, and soft smiles, creating a friendly and bright atmosphere. The woman on the left has loose long wavy hair, while the woman on the right has a high bun; both are looking naturally at the camera.

Clothing/Pose: The woman on the left is wearing {argument name="left woman's outfit" default="a slightly V-neck white ruffled blouse and light beige high-waisted shorts"}. The woman on the right is wearing {argument name="right woman's outfit" default="a pale yellow floral camisole and a soft-colored mini skirt"}. Both wear delicate accessories and display a natural, elegant S-curve silhouette. They are leaning their shoulders together, with the woman on the left making a peace sign.

Background/Light: A Japanese-style interior with shoji screens, wooden floors, modern-traditional shelving, bamboo, small flowers, soft andon-style lights, and a low round table. The walls feature hand-drawn hearts, flowers, and friendly Japanese messages, all enveloped in warm, orange-toned indoor lighting.

Composition/Camera: High-angle smartphone wide-angle selfie style looking down slightly, a full-body vertical composition with the two centered. An intimate distance that makes their faces look prominent, natural perspective, and soft background blur.

Texture/Style: High-definition 4K quality with natural skin texture visible down to pores and grain, bright and cute Japanese-style portrait, soft warm tones, pop and approachable photo expression for social media, fine hair details, and a clean finish.

Negative: low quality, unnatural face, unnatural hands, extra fingers, missing fingers, distorted fingers, strange gaze, asymmetrical eyes, collapsed face, distorted body, excessively exaggerated physique, unnatural waist curve, broken clothes, excessive processing, excessive skin retouching, excessive exposure, unnatural AI texture, distorted text, logo, watermark

A young adult Eastern female over {argument name="age" default="20 years old"}, in a portrait style combining realism with light sci-fi, with a dreamy, cold, exquisite, and futuristic temperament. She has distinctive {argument name="hair color" default="laser/holographic"} long hair with a colorful gradient effect in shades of light purple, ice blue, pale pink, and pearl, with a holographic iridescent sheen that reflects metallic colors under light.

Her features are delicate and soft with clear Eastern characteristics, a small smooth face, natural nose, full lips, and fair, translucent skin. Her gaze is quiet and clear. The outfit is a futuristic translucent holographic dress made of transparent iridescent material. The top is a sleeveless slim-cut design with semi-transparent structural stitching, while the bottom is a short pleated skirt that reflects a rainbow of colors. The overall look is fashionable, light, and dreamy with a strong futuristic tech-couture feel.

Her figure is tall and well-proportioned with elegant posture. The focus is on the hair texture, futuristic fabric, and the overall dreamy sci-fi atmosphere.

A live broadcast scene from a hot spring inn introduction program. In a {argument name="location" default="calm Japanese-style luxury hot spring inn's open-air bath"}, {argument name="presenter" default="a young and lovely adult female announcer"} is giving a bathing report. Her expression is bright and elegant, smiling at the camera as she introduces the charms of the hot spring. The background includes a wooden inn, a rock bath, bamboo fences, soft Japanese-style lighting, rising steam, and quiet natural scenery. The announcer is immersed in the water up to her shoulders, with modest exposure, a clean feel, and a stable composition typical of a travel program. The camera is slightly frontal and steady, like a TV broadcast. The overall atmosphere is bright, high-quality, and soothing. At this point, pandas have not appeared yet. Realistic, high quality, natural lighting, TV program style, elegant, and friendly atmosphere.

Referencing images 1-9. Generate a realistic and pictorial portrait of a person set in a cozy and artistic cafe. The person is painting a large-scale vertical mural portrait in a professional editorial style. They are focused on painting the mural and not looking at the camera. The person is wearing {argument name="outfit" default="jeans and a white tank top"} and a {argument name="apron" default="beige Carhartt apron"}. The subject is crouching next to a wall with a rough texture, painting a handmade mural that captures their own face, expression, and reflected visual concept. Next to the portrait, the phrase "{argument name="text" default="Love you"}" is written in a cute font with a black border. 8K ultra-high resolution detail, voluminous lighting, and professional aesthetics. Output in a professional editorial photography style while maintaining original proportions, skin tone, and expressions. Paint cans are scattered on the floor, and there are paint marks on her hands and clothes. Fact: Realistic human proportions, plastic materials prohibited.

Soft-focus dreamy portrait photography style, low contrast, slightly overexposed, airy atomization effect, the image has a hazy filter and film feel. Natural scattered light, backlight or side backlight, soft halo on the edges of the character, fair and transparent skin tone, the overall color tone is {argument name="color tone" default="light, low saturation, biased towards milky white and light green"}, the atmosphere is gentle, quiet, romantic, and girly. Shallow depth of field background blur, details are not overly sharp, like soft portrait photos taken with early digital cameras or mobile phones. The composition is mainly {argument name="composition" default="vertical half-length or seven-point body portrait"}, the character is centered or slightly to one side, the expression is {argument name="expression" default="cold, gentle, slightly melancholy"}, and the eyes look at the camera. Add elements such as soft reflections, the overall presentation is dreamy, light, ethereal, and private portrait-like photography texture. Real photography, not illustration, not 3D, not over-retouched.

2x2 grid, {argument name="source object" default="b2 plane, moon landing scene"}, 2 other complex scenes with 4 fun materials -> translation_solve(source_object → target_material): source = [object] target_mat = [material] mapping_rules: - source.primary_function → target_mat.equivalent_function_in_new_material - source.surface_detail → target_mat.characteristic_texture * 2.0 - source.joinery_method → infer(connection_type from target_mat.physics) - source.weight_distribution → target_mat.density_logic * 1.5 - source.color_scheme → infer(natural_palette from target_mat.typical_hues) render( subject = source reconstructed_in target_mat, realism = "material properties must be physically plausible" * 1.8, detail = "zoomed macro view showing material authenticity" * 1.6, context = "isolated on complementary background" * 1.2 ) + studio_product_photography - (photoshop_texture_overlay / 5) - (impossible_material_behavior / 4) example: "{argument name="example source" default="sports car"} → {argument name="example material" default="woven rattan"}"

A relaxing and quiet evening riverside scene: {argument name="time" default="late summer evening"}, {argument name="setting" default="at a quiet riverside walkway on the edge of the city"}, the wind is very light, the sky is a transition moment of gray-blue with a bit of warmth. A {argument name="character" default="young Oriental girl"} is sitting on a low wall by the river, wearing a simple light-colored knitted cardigan and a dark skirt, white canvas shoes landing randomly, her posture natural and relaxed, without a clear posed feel. She is holding a small bouquet of freshly bought flowers, with a canvas bag beside her. The river surface is calm, the leaves on the shore are gently blown by the wind, and the opposite bank is blurred low buildings and scattered warm lights. The whole picture is clean, comfortable, and restrained, focusing on the evening breeze, quiet water surface, clothing texture, skin tone, tree shadows, and twilight air. Colors are mainly gray-blue, light beige, soft dark green, and small warm yellow lights, not intense or theatrical, like a real and slightly poetic film photograph. photorealistic, calm riverside evening, soft ambient dusk light, natural candid pose, subtle film-like color grading, breathable composition, quiet emotional realism, no text, no watermark

An authentic Korean-style Instagram bathroom mirror selfie, shot on a smartphone by a {argument name="age" default="25-year-old"} East Asian woman inside a luxury hotel suite. Her figure is mature and naturally hourglass with healthy, realistic proportions: soft fuller upper-body curves, a naturally narrow waist, and an overall rounded feminine silhouette. She carries the styled aura of a Korean influencer, {argument name="hair style" default="long voluminous wavy brown hair flowing naturally"}, subtle refined makeup, fair clear skin with a natural glow, gentle features, naturally rosy lips, eyes softly lowered toward the phone screen, a calm mature expression. She wears a {argument name="outfit" default="minimal chic triangle swimsuit in a black-and-white floral print"} that follows her natural curves with soft fabric tension and drape. She sits cross-legged on the clean floor of a modern luxury hotel bathroom, casually snapping the selfie. One leg tilts slightly toward the camera as a soft foreground, body angled toward the mirror, one hand holding a black phone that partially covers her face, the other hand resting naturally on the floor, a real lived-in relaxed post-vacation pose. Setting: a minimalist luxury hotel bathroom with beige tiled walls, a black vanity cabinet, a dark wooden door, warm soft interior lighting, low contrast, and a natural sense of space. Vertical 3:4 framing, close-range phone-selfie viewpoint, slight natural wide-angle distortion, subject filling the main frame, mirror-reflection composition. Overall an authentic phone-shot texture: Korean Instagram-selfie aesthetic, lightly edited in social-media style yet keeping real skin texture, soft focus, mild film grain, natural warm light and shadow, no over-smoothing, no plastic skin, a strong life-record quality with a refined real luxury daily mood. Real human proportions, naturally mature feminine beauty, luxury hotel staycation life-record style.

Generate a 16:9 horizontal infographic poster. Theme: Pan Jinlian. Must cover knowledge points: [Point 1] through [Point 8]. Visual Style Definition: This is a structural flat infographic poster in the style of 'Perler beads / fuse beads / pixel bead array.' The image consists of a large number of regularly arranged circular plastic beads, each with clear circular boundaries, a slight central indentation, uniform spacing, low-saturation plastic texture, and stable grid order. The overall composition must maintain a front-facing flat view, resembling a public cultural infographic poster made of beads, rather than toy photography, 3D models, cartoon illustrations, or ordinary pixel art. Core Composition: Compress the [Theme] subject into a large-scale monochromatic bead image field rather than a complete realistic object or centered illustration. The subject must flow in from the edges of the frame, crossing page boundaries with a bleed cut, like a larger fragment of bead media entering the page. The subject occupies the main visual weight but is not fully presented; viewers must reconstruct the theme through silhouettes, directions, missing areas, bead density, and local textures. Subject Morphology: The outer edges of the subject show low-resolution bead steps, block fractures, hard-cut gaps, pixelated jagged outlines, missing bead sections, and rough boundaries after sampling. Smooth silhouettes, complete icon outlines, and realistic details are prohibited. The interior uses only different brightness levels of the same thematic structural color to form low-contrast image fragments, halftone noise, scanning grains, material afterimages, archival textures, and local density changes. Background and Negative Space: The background uses a high-brightness light-colored bead field, similar to unprinted paper, light creamy white plastic baseboards, or pale grey-white bead arrays. The background is not a decorative base color but actively cuts into the subject, forming large voids, winding paths, quiet text windows, reading pause zones, and knowledge point containers. Whitespace must act like air and a blade; it provides breathing room and cuts the subject structure. Large light-colored areas and large structural color areas must invade each other. Color System: Adopt a strict three-layer functional color scheme. Layer 1: Light-colored base (60%-70% area) for breathing and windows. Layer 2: Theme structural color (25%-35% area) for the subject and mood. Layer 3: High-contrast information color (3%-6% area) for titles, numbers, and annotations. Infographic System: Convert all 8 knowledge points into bead information nodes using forms like numbered tags, coordinate lines, or time scales. Reading Path: Build a jumping reading path from top markers to main titles to peripheral knowledge points and finally to the footer. Typography: Use narrow modern sans-serif fonts with precise hierarchy. Theme Translation Rules: Do not use default icons; translate concepts into bead density, missing areas, or spatial pressure. Material: All elements are made of beads with matte plastic textures and slight manufacturing irregularities. Visual Ambience: Restrained, rational, modern, and archival. Hard Requirements: 16:9, flat view, circular bead particles clearly visible, subject bleed-cut, at least 8 information nodes. Forbidden: Realistic objects, centered illustrations, cartoon styles, 3D toy photography, depth of field, glows, or gradients. Final Effect: A high-end structural infographic poster where the theme is represented by bead fields and pixelated voids rather than being directly drawn.

Use only my uploaded portrait as the identity reference. Strictly preserve my real facial features, proportions, face shape, hairstyle, hairline, skin tone, and overall vibe. Do not use the reference image’s face, gender traits, or facial appearance. Only use the reference image for the camera language, pose, scene, composition, lighting, and atmosphere. Create an ultra-realistic urban street photo: the person is sitting on the ground or steps at the entrance of a modern city building, photographed from a very low angle with an ultra-wide fisheye lens. The camera is extremely close to the ground, creating strong curved spatial distortion. The person is positioned near the lower center of the frame, with one hand naturally reaching toward the camera, the palm very close to the lens to create a dramatic foreground perspective. The other leg is bent naturally. The pose should feel relaxed and natural, with a real, soft smile. The setting is a modern city street surrounded by tall buildings, glass doors, steps, trees, and distant architecture, all slightly stretched by the fisheye effect. Golden-hour sunlight shines from behind the subject, creating natural backlight, a sun flare, and long shadows. The overall lighting is warm and cinematic. The image should look like a real action-camera or ultra-wide smartphone photo, with realistic skin texture, natural pores, believable shadows, and well-integrated environmental lighting. Do not over-smooth the skin. No plastic look, no anime look. The outfit should be casual summer streetwear, automatically styled to suit my appearance, such as a {argument name="clothing item" default="simple white T-shirt"}, {argument name="bottoms" default="shorts"}, and sneakers. Keep it natural and not overly styled. The final image should feel realistic, candid, energetic, and documentary-like, in a vertical 4:5 composition, with no watermark, no text, and no logo. Negative Prompt: watermark, text, logo, AI label, anime style, cartoon style, plastic skin, over-smoothed skin, fake face, face-swap look, poor likeness, distorted facial features, bad hands, extra fingers, deformed palm, twisted limbs, low resolution, blur, overexposure, oversharpening, unrealistic lighting, pasted-on subject, studio portrait look, heavy makeup, incorrect perspective, messy background

Please generate an ultra-high quality storyboard sheet for a 15-second anime cooking video as a single landscape 16:9 image. The theme is "{argument name="theme" default="rich cheesy omelet rice"}". Within a single landscape canvas, arrange a 12-frame storyboard in chronological order. The layout should be a 4-column by 3-row grid.

Goal: Create a sleek monochrome brand guidelines cover for {argument name="brand name" default="UtopikAI Studio"}, presenting it as a premium AI studio visual identity system.

Canvas: Square 1:1 format, black charcoal background with subtle grain, faint vignette, and thin technical drafting lines. Use a minimal futuristic editorial layout with generous negative space.

Layout: Add a very thin rectangular border close to the canvas edges, with tiny corner ticks, small crosshair marks at the top center, left middle, right middle, and bottom center, and faint dotted vertical guide marks near both side edges. In the upper left, place small spaced uppercase text “UTOPIKAI STUDIO”; in the upper right, place “BRAND GUIDELINES”. In the center, build a large geometric construction grid: a 2-by-2 square grid with faint circular arc segments and quarter-circle curves, like logo construction guides. Overlay the main logo lockup across the middle-lower portion of this grid.

Logo lockup: Use a white abstract icon to the left, made from four square tiles containing curved quarter-circle cutouts, forming a symmetrical modern mark. To the right, set the wordmark “UtopikAI” in bold clean geometric sans serif, bright white, with “Studio” smaller and lighter aligned under the right side of the wordmark. Count exactly 2 logo elements: 1 abstract icon and 1 two-line wordmark.

Main title area: In the lower left quadrant, place the large title {argument name="document title" default="Brand Kit"} in white, set in a clean modern sans serif. Directly beneath it, add the subtitle {argument name="subtitle text" default="Visual identity system"} in smaller muted gray text. Under the subtitle, add one short thin horizontal white line and one longer very faint gray line.

Footer: Along the bottom edge, add small technical footer labels: left “VERSION 1.0”, center “VISUAL IDENTITY SYSTEM”, and right {argument name="year" default="2024"}. Connect the footer area with thin angular drafting lines and subtle horizontal rules.

Visual style: Premium black-and-white brand kit cover, minimalist, tech-luxury, precise Swiss grid design, high contrast white typography, subtle gray construction lines, crisp vector-like rendering, no mockup perspective.

Constraints: Use exactly 5 main text groups: top-left studio label, top-right guidelines label, central logo lockup, lower-left title/subtitle block, and bottom footer metadata. Keep all text sharp and legible. Avoid extra colors, photos, people, shadows, or decorative clutter.

Create a wide hand-drawn editorial illustration of a surreal professional restaurant kitchen for {argument name="project name" default="a cursed repo visualization"}. Use a 16:9 canvas with a slightly elevated three-quarter view down a long white marble prep counter. The scene should feel like a whimsical, cursed software-repository metaphor rendered as a detailed ink-and-watercolor kitchen: warm cream walls, dark cabinetry, copper pans hanging along the back wall, brass pendant lamps, recipe tickets clipped to rails, towels on hooks, a chalkboard menu with illegible scribbles, stacks of plates, folded napkins, knives, tongs, saucepans, steaming bowls, and many tiny handwritten ingredient labels with mostly nonsensical text. Include exactly 5 human chefs: 1 dim ghostlike chef on the far left fading into a dark vignette, 1 chef at the left-center counter plating food, 1 partially obscured chef near the back center leaning over a pot, 1 large main chef seen from behind in the foreground wearing a tall toque and black-and-white checkered pants, and 1 chef on the right actively stirring a pan. Add exactly 1 anthropomorphic shrimp chef mascot standing on a small wooden step stool at the lower right, wearing a white chef coat and tall chef hat, holding a fan of pink order tickets like playing cards. The counter should contain exactly 12 prominent plated or bowled food items: 1 covered silver cloche near the lower left, 1 bowl of pale round beans labeled like gibberish, 1 bowl of mixed diced ingredients, 1 white plate of golden balls, 1 small red sauce bowl, 1 copper pan of round items steaming near the center, 1 plated tiny arrangement being assembled by the left-center chef, 1 central lidded or domed dish, 1 shallow bowl of grain-like filling, 1 reddish stew bowl, 1 white plate with a tiny plated garnish on the right, and 1 white plate of browned round bites at the far right. Scatter many small paper labels in front of the dishes; make the labels look hand-lettered and intentionally wrong or hard to read, including fragments resembling “BRUNCISE,” “JULIENNE,” “CANDI HANGER,” and other pseudo culinary/software terms. The left third of the image should be covered by a heavy dark smoky shadow and the right side should be clearer and warmer, creating an ominous-to-comedic contrast. Style: delicate black linework, muted watercolor wash, slightly vintage cookbook illustration, high detail, soft steam, no photorealism, no modern appliances, no clean vector graphics. Keep all text small, imperfect, and mostly illegible except for a few goofy labels. Use {argument name="overall mood" default="cursed but charming"} as the emotional tone, {argument name="mascot species" default="shrimp"} for the tiny chef mascot, and {argument name="color palette" default="warm cream, copper, charcoal, white marble, muted red accents"}.

Using the provided reference image as the izakaya interior mood reference, transform the wall menu into a closer, more frontal handheld snapshot of a casual Japanese pub menu corner. Replace the original white printed paper notices on the wooden wall with exactly 12 narrow cream menu slips outlined in red, arranged in 2 rows of 6, each pinned at the top and written in bold black brush-style Japanese with orange handwritten prices. The 12 menu slips should read: top row 「だしまき 580」「モロキュー 400」「串カツ 150」「豚キムチ 580」「もずく酢 380」「冷やっこ 300」; bottom row 「牛すじポン酢 680」「たこぶつ 680」「チャンジャ 480」「鶏の唐揚 680」「ウインナー 480」「エイヒレ 530」. Add a black chalkboard menu above the slips with messy white chalk writing and a few orange prices, plus a yellow beer poster on the left wall. Keep the warm dim izakaya lighting and lived-in, slightly blurry phone-photo realism, but shift the background from a full wooden plank wall to a cream wall with dark wood wainscoting at the bottom. Include small red-capped condiment bottles partially visible at the bottom edge. Optional text customization: use {argument name="top row menu text" default="だしまき 580, モロキュー 400, 串カツ 150, 豚キムチ 580, もずく酢 380, 冷やっこ 300"} and {argument name="bottom row menu text" default="牛すじポン酢 680, たこぶつ 680, チャンジャ 480, 鶏の唐揚 680, ウインナー 480, エイヒレ 530"}.

Create a candid photorealistic film-style portrait of a young woman named {argument name="character name" default="anonymous young woman"} standing on a sunny waterfront promenade in Yokohama, Japan, with the Minato Mirai skyline in the background. She is shown from the waist up in a relaxed three-quarter back pose, turning slightly toward the camera, one hand raised near her hair, with her face intentionally covered by a plain soft-edged rectangular privacy blur in a neutral skin-tone/brown color. She has very long flowing {argument name="hair color" default="black"} hair falling down her back, pale skin, and wears exactly 3 visible accessories on her wrist: a gold bracelet, a beaded bracelet, and a thin chain bracelet. Her outfit has exactly 1 visible main clothing piece: an oversized off-shoulder {argument name="shirt color" default="white"} T-shirt with a faint gray graphic print on the lower front. The setting includes exactly 5 key background landmarks or structures: a tall rectangular Landmark Tower on the left, several pale modern high-rise buildings in the center, a large Ferris wheel on the right, a waterfront railing with round metal posts along the right edge, and calm blue harbor water beside the promenade. Use natural morning sunlight, a gentle sea-breeze feeling, soft shadows, muted pastel colors, shallow depth of field, slight 35mm film grain, realistic proportions, and a vertical 3:4 composition. No text, no watermark, no extra people in the foreground.

Create a highly detailed 4K vertical anime-style illustration of {argument name="character description" default="a young Japanese woman"} standing in a lush traditional Japanese garden just after rain, surrounded by bright fresh green maple leaves. She wears {argument name="kimono design" default="a white yukata kimono patterned with green maple leaves and small pale blue flowers"}, tied with {argument name="obi color" default="an olive-green obi with a small floral ornament"}, and green-and-white geta sandals. Her dark hair is pinned up in a soft elegant bun with loose strands and small flower hair ornaments; her face is intentionally covered by a smooth rectangular skin-toned blur/blank censor block with no facial features. Pose her leaning forward slightly from the garden path, one arm raised to gently lift a low maple branch above her head, the other hand holding a small pale green drawstring pouch. The setting is a narrow wet stepping-stone path with puddles and glossy reflections, mossy rocks, low shrubs, hydrangea-like blue flowers, a bamboo fence and open wooden gate on the left, a bamboo fence in the background, and a weathered stone lantern on the right. Use soft morning sunlight filtering through dense maple foliage, sparkling raindrops, vivid greens, warm backlight, painterly realism, delicate linework, cinematic depth, rich environmental detail, tranquil romantic atmosphere, no text, no watermark.

Create a cinematic wide anime illustration of two futuristic schoolgirl idols seen from behind, holding hands while running or floating across a mirror-like shallow ocean at sunset beneath a vast starry sky. The left girl has {argument name="left character hair color" default="long pink twin-tail hair"}, a black oversized streetwear jacket with punk straps and graphic panels, a short pleated skirt, thigh straps, chunky lace-up platform sneakers, and large futuristic headphones with a bow-like accessory; she is backlit by a warm pink sunset and framed by dense cherry blossom branches on the far left. The right girl has {argument name="right character hair color" default="dark blue twin-tail hair"}, a white oversized jacket with black abstract patterns, a dark pleated skirt, thigh straps, chunky blue-and-white sneakers, and a large ribbon accessory in her hair; her side is lit by cool cyan moonlight and blue sky tones. Their arms stretch toward the center and their fingers interlace delicately, with petals and tiny light particles swirling around them. The background is split into a dramatic pink-to-blue gradient: warm rose sky and glowing low sun near the left horizon, transitioning into deep blue night sky with stars, nebula-like clouds, shooting stars, and a crescent moon on the right. The water surface reflects both figures, cherry blossoms, sunset colors, and blue sky with glossy ripples. Use a highly detailed Niji-style anime aesthetic, vibrant saturated colors, dynamic wide-angle composition, elegant linework, luminous rim lighting, dreamy magical atmosphere, intricate clothing details, flowing hair ribbons, and abundant sakura petals. Keep exactly two girls as the main subjects, viewed mostly from behind, no text, no watermark, no extra characters, high-resolution 16:9 landscape composition, polished gpt-image-2 anime refinement quality. Optional mood: {argument name="mood" default="bittersweet magical spring night"}.

Create a richly detailed whimsical storybook illustration of a winter night lantern festival in a cozy village, in the style of a hand-painted children’s picture book with gouache and colored-pencil texture. Set the scene under a deep navy star-filled sky with glowing circular stars and tiny snow-like speckles. The ground is pale icy blue snow, framed by dense decorative shrubs and tall coral, teal, blue, orange, and yellow trees. Use a wide cinematic canvas, approximately 16:9.

Scene: A small village prepares for {argument name="festival theme" default="a glowing paper lantern festival"}. Show exactly 5 warm wooden cottages with dark blue tiled roofs: 1 large cropped cottage on the far left, 1 small lit cottage left of center, 1 small cottage in the middle background, 1 cottage right of center with round attic window, and 1 large cropped cottage on the far right decorated with garlands. All cottage windows glow warm yellow-orange.

Main layout: Place a tall central wooden pole or A-frame structure slightly left of center, wrapped with golden garlands and topped with a red starburst ornament. Hang strings of triangular bunting in orange, blue, red, and yellow across the sky. Add glowing paper lanterns in red, orange, yellow, and cream throughout the scene, both hanging and being carried. Include exactly 3 ladders: 1 ladder on the far left with a villager hanging a lantern, 1 tall central ladder/A-frame supporting decorations, and 1 ladder leaning against the right cottage with a villager decorating.

Characters and counts: Include exactly 18 small stylized villagers with childlike proportions, round faces, simple dot eyes, rosy cheeks, and winter clothing in deep blue, red, green, orange, brown, and cream. Count and place them as follows: 1 person on the far-left ladder holding a red lantern; 1 child near the central ladder holding an orange lantern; 2 villagers left foreground exchanging a red lantern; 1 teal-coated villager and 1 small orange-haired child near the left path; 1 red-coated child in the lower-left foreground holding a lantern; 2 villagers at the central wooden worktable arranging lanterns; 1 blue-coated woman near center carrying a lantern; 1 orange-coated man near a flower stand; 1 dark-blue vendor behind a small display table; 2 small villagers in the right background walking away; 1 woman in an apron holding branches; 1 small green-coated child holding a basket; 1 orange-coated woman carrying branches; and 1 witch-hat-like dark-blue figure on the far right hanging garlands. Add exactly 1 black cat sitting near the central-left work area.

Props: Include exactly 1 central wooden worktable covered with paper lanterns and floral sprigs, exactly 3 wooden crates filled with lanterns and flowers around it, exactly 1 small flower display stand in the mid-right background, exactly 2 large barrels or tubs with branches near the right cottage, and scattered baskets of flowers and branches. Make the lanterns glow softly and cast warm highlights on nearby faces, snow, and wood.

Visual style: Use a magical, cozy, highly detailed folk-art aesthetic with flattened perspective, delicate linework, dense ornamental foliage, soft warm window light, and a peaceful festive mood. The palette should contrast cool midnight blues and icy snow with warm lantern oranges, reds, and yellows. No text, no captions, no watermark, no modern objects, no photorealism.

Create a cinematic before-and-after transformation diptych featuring an adult woman in ancient Chinese fantasy styling. Canvas: horizontal 4:3 image split vertically into exactly 2 equal panels with a clean center seam. Left panel: write the pink handwritten label “Before” in the upper left; show the woman seated in a dark wooden pavilion, wearing a flowing white hanfu with translucent veils, gold embroidery, and an elegant silver hair crown, with long black hair and a decorative sword leaning diagonally along her left side; the mood is solemn wuxia fantasy, with mountains and traditional architecture softly visible in the background. Right panel: write the pink handwritten label “After” in the upper right; show the same woman transformed into a softer pink floral hanfu look with a pink choker, layered robe, embroidered sash, flower hair ornaments, and long black hair styled with loose volume; she poses playfully with both hands making V-signs near her cheeks. Add exactly 1 thick curved pink arrow across the top center pointing from the left panel to the right panel. Cover the face in each panel with an opaque rectangular censor block in muted brown, exactly 2 censor rectangles total, while keeping the hair, outfit, pose, and body visible. Visual style: realistic high-detail cosplay photography, warm indoor sunlight, shallow depth of field, soft filmic color grading, ornate historical Chinese room details, polished fantasy costume textures, no extra text besides the two labels. Customizable subject details: {argument name="character name" default="ancient fantasy heroine"}, {argument name="before outfit color" default="white"}, {argument name="after outfit color" default="pink"}, {argument name="before label" default="Before"}, {argument name="after label" default="After"}.

Goal: Create a side-by-side comparison graphic demonstrating global illumination / indirect bounce lighting improvement on the same anime illustration.

Canvas: Wide 16:9 horizontal image, split exactly into two equal vertical panels with a thin white divider line down the center. Use a soft pastel anime illustration style with dreamy morning light.

Layout: The left panel is labeled “Before” at the top-left and shows the normal version; the right panel is labeled “After” at the top-right and shows the enhanced version with warmer indirect light, natural contact shadows, and soft glow wrapping around the subject. Add a large centered headline across the top reading {argument name="headline text" default="プロンプト比較：光の回り込みGI"}. Add bottom-left Japanese label {argument name="left bottom label" default="通常生成"} on the left panel and bottom-right Japanese label {argument name="right bottom label" default="光の回り込みGI"} on the right panel. All overlay text is bold white with a dark drop shadow for readability.

Subject details: Show the same androgynous anime boy in both panels, seated indoors in a cozy bedroom by a bright window. He has fluffy layered {argument name="hair color" default="white-silver"} hair with lavender shadows, pale skin, and a slim elegant build. His face is intentionally covered by one large opaque pale mauve square censor block centered over the face in each panel; use exactly 2 censor blocks total, one per panel. His right hand rests near his cheek, partially visible beside the censor block. He wears a luxurious white robe or kimono-like top with navy-blue inner layers and wide sleeves, trimmed with metallic silver edges, decorated with star and crescent moon motifs. Include a crescent moon pendant necklace with a dangling blue gemstone.

Background details: The room contains exactly 1 large window visible behind the subject, soft curtains, a bed or cushioned seat, a small shelf or windowsill with perfume bottles and ornaments, and exactly 1 hanging crescent-moon decoration on each panel. The palette is lavender, white, pale blue, and warm sunlight.

Lighting comparison: Left “Before” panel should look flatter and cooler, with less bounce light and weaker shadows. Right “After” panel should visibly improve depth: add soft global illumination, warm sunlight wrapping around hair, face area, hands, robe folds, and shoulders; add subtle contact shadows where the hand touches the face area and where sleeves overlap; add bloom near the window and brighter rim light on the hair. Keep the character pose and composition identical between panels.

Constraints: Countable elements must remain exact: 2 comparison panels, 1 vertical divider, 4 text labels total, 2 face censor blocks, and 2 hanging crescent decorations. No watermark, no extra logos, no additional panels.

Create a surreal high-fashion studio editorial photograph featuring {argument name="model description" default="a glamorous young woman with long dark wavy hair and warm medium skin"} seated on a plush pink faux-fur armchair in an all-pink fur-lined set. She wears an oversized voluminous {argument name="coat color and material" default="bubblegum pink faux-fur coat"} draped around her body, bare legs crossed elegantly, and sparkling pink pointed stiletto heels. She holds a clear glass bottle filled with milk in both hands at her lap and looks calmly into the camera with a poised fashion-model expression. Beside her, a {argument name="animal model" default="cream-colored calf"} leans in from the right, resting its head affectionately on the chair arm and close to her shoulder; show the animal’s full head, neck, torso, and front legs, natural and realistic. The entire environment is monochrome pink: fuzzy wall backdrop, fuzzy floor, and rounded furry chair, with rich tactile texture everywhere. Use glossy magazine fashion photography styling, surreal luxury dairy concept, soft yet directional studio lighting from upper left, gentle shadows, warm highlights on skin and fur, shallow depth of field, realistic proportions, high detail, shot on a {argument name="lens" default="50mm fashion editorial lens"}, square 1:1 crop, centered composition, no text, no logos, no watermark.

Goal: Create a dramatic two-panel Japanese fantasy manga comparison comic about how large damage numbers make combat feel more satisfying, even when the time required is the same.

Canvas: Square 1:1 image, full-color manga illustration with thick black comic borders, dynamic speed lines, explosive impact effects, and a black brushstroke title banner at the top. Overall style should be detailed dark-fantasy shonen manga with bold outlined lettering.

Layout: Use exactly 2 wide horizontal comic panels stacked vertically. Add a black brushstroke header above both panels containing the title text {argument name="title text" default="『かかる時間は同じでも……』"}. Each panel has the same party of 3 heroes on the left and 1 enemy on the right, with a vertical white narration box on the far right.

Characters: The hero party appears in both panels and contains exactly 3 visible characters: 1 young male swordsman in blue armor holding a sword, 1 young female witch in a purple robe and pointed hat, and 1 small green-haired healer or cleric holding a staff. Their faces may be partially obscured or softly blurred, but their poses and outfits remain readable. Top panel enemy: 1 enormous black armored knight/golem with a shield and heavy plate armor, standing on the right. Bottom panel enemy: 1 enormous bare-chested muscular barbarian/giant with rough skin, leather/fur waist gear, and clenched fists, standing on the right.

Top panel text and effects: Include exactly 2 white speech bubbles from the heroes: 「これも効かないの!?」 and 「どうしたらいいんだよコイツ!」. Include exactly 1 vertical narration box on the right reading 「耐性がカッチカチ、よりも」. Add exactly 4 small blue damage numbers around the armored enemy: 「10!」, 「10!」, 「12!」, and 「12!」.

Bottom panel text and effects: Include exactly 2 white speech bubbles from the heroes: 「こいつ、まだ立ち向かってくる……!」 and 「なんてタフなんだ!」. Include exactly 1 vertical narration box on the right reading 「HPモリモリの方が」. Add exactly 2 huge fiery red-orange damage numbers: 「1012!」 near the center-left and 「2001!」 near the right. At the bottom, spanning across the lower panel, place a large energetic slogan in thick yellow-orange Japanese brush lettering with black outline: {argument name="bottom slogan" default="数字がでっかくて楽しい!"}.

Visual style: High contrast, gritty fantasy lighting, dramatic explosions behind enemies, manga hatching, saturated impact colors, thick white speech bubbles with black outlines, bold 3D comic typography for numbers. Make the top panel feel frustrating and low-impact with small blue numbers; make the bottom panel feel exciting and powerful with huge warm-colored numbers.

Constraints: Use exactly 2 panels, exactly 4 total main characters per panel, exactly 4 small damage numbers in the top panel, exactly 2 large damage numbers in the bottom panel, exactly 4 total speech bubbles, exactly 2 vertical narration boxes, and 1 large bottom slogan. Keep all visible Japanese text legible and avoid adding extra text or watermarks.

Create a polished product mockup showing a cozy winter campfire wallpaper adapted across three Apple-style devices. The wallpaper illustration should resemble a cute Japanese crayon-anime style without logos: five small cartoon children and one small white dog gathered around a bright campfire in a snowy pine forest at night, with falling snow, deep navy-blue trees, warm orange firelight illuminating the characters and snow, and a soft magical winter atmosphere. The main character is a round-faced child in a red winter sweater standing behind the fire, surrounded by four friends: one with dark spiky hair on the left, one in a brown coat at lower left, one curly-haired child in pink on the right, and one child in a green hood at lower right; the white dog sleeps or curls up near the fire in front. Show exactly three devices on a light gray studio background, each inside a rounded white display card with soft shadows: one large laptop mockup centered at the top, one tablet mockup at bottom left, and one smartphone mockup at bottom right. On all screens, display the same wallpaper composition, cropped appropriately. Add realistic lock-screen UI overlays: large time {argument name="time" default="9:38"}, date {argument name="date" default="Friday, November 15"}, small status icons, and subtle translucent widgets. The tablet should include a translucent weather widget on the upper left showing {argument name="weather location" default="Cupertino"} and {argument name="temperature" default="45°"}; the phone should include three circular widgets below the time and two small bottom lock-screen icons. Use clean high-end Apple product rendering, black bezels, thin device frames, rounded corners, soft ambient shadows, sharp screen detail, and a balanced wallpaper-preview layout suitable for desktop, tablet, and mobile backgrounds. Avoid brand logos, watermarks, extra text, or additional devices.