GPT Image 2Prompts

Using cinematic quality, Kodak film style, slightly reduced saturation, high dynamic range, and realistic contrast. Lens: {argument name="lens" default="85mm lens"}, {argument name="aperture" default="f/1.8 aperture"}, shallow depth of field, RAW format photorealistic feel, rich detail. Generate a cinematic character poster featuring the {argument name="character name" default="character's name"}. It's both a character poster and a representation of their story. A large, elegant profile of the character's face occupies half the image, with perspective and movement, creating a slightly wide-angle lens feel. The character's appearance and clothing match the original character design. Skin pores. Correct anatomy. Fingers relaxed naturally. No gloves. Facial and clothing features should also be adapted from the original design, including the weapon held in the hand, and expressions and poses consistent with the character design, generating a 9:16 poster image.

Goal: Create a highly detailed vertical photorealistic miniature diorama inside a clear glass display case, styled like a warm pastel Japanese kawaii room box, with the feeling of a cozy personal world. Canvas: Portrait 2:3 composition, close-up front view through transparent glass, shallow depth of field, warm golden afternoon light, soft bokeh fairy lights, ultra-detailed handcrafted textures, realistic reflections on the acrylic/glass walls. Main subject: A seated teenage girl miniature with a softly obscured/featureless face, shoulder-length {argument name="hair color" default="dark brown"} hair, wearing a cream Japanese sailor school uniform with gray collar trim, a large {argument name="bow color" default="pink"} chest bow, gray pleated skirt, white socks, sitting sideways on a cream lace-covered armchair. Her left arm reaches outward holding a tiny pink flower or wand-like blossom; her right hand rests on her knee. Pose is gentle, relaxed, and inviting. Companion subject: One small floating fairy-like robot girl to the right of the seated girl, also with a softly obscured/featureless face, short brown bob hair, translucent insect wings, white-and-silver futuristic body armor, tiny mechanical limbs, hovering above a small round flying base. Make it cute, delicate, and toy-like rather than threatening. Display case and room: The scene is enclosed in a rectangular glass case with a wooden base and ornate gold vine decoration along the bottom. Inside is a miniature living room with a sofa/armchair, lace curtains, sunny window, bookshelves, framed photos, houseplants, cushions, a small wooden coffee table, pastel boxes, a cup and saucer, tiny cosmetics or perfume bottles, and a soft cream rug. Use peach, ivory, blush pink, champagne gold, and pale green as the main palette. Decorative arch: At the top front of the case is a floral arch made of vines, pearls, crystal drops, warm micro string lights, and many small white and pink flowers. Add one ornate oval sign centered at the top reading {argument name="top sign text" default="まるこのおうち時間"}, with a small cursive English subtitle reading "Maruko's little world". Add one matching name plaque near the bottom front reading {argument name="name plaque text" default="まるこ"} with "Maruko" underneath. Counted visible focal elements: Include exactly 2 human-like figures: 1 seated schoolgirl miniature and 1 floating winged robot fairy. Include exactly 1 sleeping white cat curled in a small woven basket on the floor near the girl's feet. Include exactly 2 readable signs: the top oval title sign and the bottom name plaque. Include exactly 1 small wooden coffee table, 1 cup-and-saucer set, 1 armchair, 1 window with lace curtains, and 1 wooden display-case base. Style and lighting: Premium collectible dollhouse photography, hyperrealistic miniature craftsmanship, romantic kawaii aesthetic, cinematic warm sunlight from the back window, glowing fairy lights, crisp foreground details with soft background blur, glass reflections and refractions, delicate lace, pearl, resin, and gold-metal textures. Constraints: Do not add extra characters, do not make it a flat illustration, do not use harsh colors, avoid horror or cyberpunk mood, keep both faces intentionally soft/blurred and non-identifiable, preserve the cute handcrafted display-case look.

A {argument name="subject" default="pretty girl"} in a crowded {argument name="setting" default="medieval market"}, market vendors, upper body, detailed clothes, {argument name="eye color" default="green"} eyes, long flowing hair, shiny skin, subsurface scattering, (sharp:0.7), ultra-detailed, real-life texture, dramatic lighting, dynamic composition, unreal engine, Nikon D850 <lora:more_details:0.6>

Goal: Create a dark cinematic AI-drama workflow showcase image: a 3×3 fantasy storyboard grid on the left and a widescreen video-preview frame on the right, demonstrating how a still storyboard becomes an animated scene. Canvas: Wide horizontal composition, roughly 2:1 aspect ratio, with a black background and thin white outer border. The left half contains a square storyboard contact sheet; the right half contains a large embedded video-player still with black letterboxing and a small playback time indicator. Left storyboard layout: Use exactly 9 numbered panels arranged in a 3×3 grid, each separated by thin white borders and marked with small white number tags from 1 to 9 in the upper-left corner. The panels should depict a gothic xianxia / dark fantasy romance narrative featuring {argument name="female character" default="a veiled bride in ornate crimson robes"} and {argument name="male character" default="a cold white-haired immortal man in dark robes"}. Add small cinematic Chinese-style subtitle text in one or two panels, but keep it tiny and secondary. The 9 panels are: 1) extreme close-up of the crimson veiled bride covering part of her face with her sleeve, red candlelight and rainlike texture; 2) the bride seated or leaning in a dim chamber near a window, a pale white-haired figure behind her, faint subtitle line near the bottom; 3) the bride turning toward a red-lit palace interior with long black hair and red veil flowing; 4) a tall dark-robed figure standing before a lattice window, backlit by cold blue moonlight; 5) the bride kneeling or collapsing beneath a red veil, one hand on the floor, dramatic red fabric; 6) intimate close-up of the white-haired immortal touching the bride’s chin, their faces close, tense romantic mood; 7) close-up of the white-haired man holding a vertical talisman or paper slip with Chinese characters; 8) two dark silhouettes facing each other in rain or mist, a red candle or ember glowing at the bottom; 9) the bride in red standing on a high balcony or rooftop, long robe and hair blown by wind, overlooking a stormy blue cityscape. Right video-preview layout: Place one large widescreen cinematic frame occupying the right half. It shows the same crimson bride from behind or three-quarter view inside a red wooden palace chamber, with black hair and a thorny dark headdress or tangled veil, red robes glittering with embroidery, blue moonlight visible through lattice windows, and tiny distant figures in the background. Add thick black letterbox space above and below the frame. In the lower-left of the black player area, show a white pause icon and the timestamp “0:13”. Include a small faint label in the top-left of the video frame, like an AI-generation tag. Visual style: Ultra-dramatic Chinese dark fantasy, gothic romance, high-contrast red and blue lighting, candlelit interiors, rain, mist, ornate robes, cinematic depth of field, moody shadows, film still texture, high detail, atmospheric fantasy drama. Use {argument name="primary color palette" default="crimson red, cold blue, black, and pale silver"}. Face treatment: Every visible human face should be deliberately obscured by a soft square gray blur, as if anonymized in a screenshot. Keep the blur blocks centered over faces and consistent across both storyboard and video frame. Text constraints: Keep all text minimal and small, except the visible timestamp “0:13” and the storyboard numbers 1 through 9. Do not add logos, watermarks, extra UI panels, or additional storyboard frames.

render_target = ( 3d_surreal_terrain_map_of_{argument name="dream theme" default="[dream_theme]"}_dreamscape * 1.2 ) + ( dream_symbol_dictionary_infographic_ui_with_{argument name="key symbol" default="[key_symbol]"} and {argument name="emotional tone" default="[emotional_tone]"}_callouts * 1.5 ) + ( [cultural_framework]_interpretation_lens_panel_with_cross-reference_index * 1.3 ) + ( lucidity_meter_widget_set_to_[lucidity_level]_with_reality_check_reminders * 0.9 ) + ( dream_journal_notebook_overlay_with_handwritten_key_moments * 0.7 ) + ( ambient_studio_background_gradient_matching_[emotional_tone]_lighting * 1.0 ) - ( generic_floating_zzz_and_sandman_dust / 3.0 ) - ( cliché_flying_through_clouds_or_teeth_falling_out_default / 2.5 ) - ( overused_brain_x-ray_and_gears / 2.0 ) - ( flat_2d_psychology_textbook_diagram / 2.5 )

Transform the attached image into a collage artwork. Make it appear as if hand-torn from newspapers, magazines, and flyers and pasted. Every single expression should be completed using large, torn pieces of paper. Represent in detail the torn edges, wrinkles, overlaps, and glue marks on the paper. Use relatively large pieces of paper, not too small, and place them randomly at different angles and directions, with the paper orientation rotated haphazardly. Create it to look like an actual collage roughly hand-pasted by a person.

Create a striking vertical poster illustration of {argument name="character description" default="a young schoolgirl in profile"} walking toward the right, rendered almost entirely as a black silhouette against a saturated blood-red background. The character has a short bob haircut blown backward by wind, a bowed head, a sailor-style school uniform with a large scarf or neckerchief trailing behind, a flared pleated skirt, long sleeves, dark stockings, and chunky shoes. Use sharp cream-white highlights along the hair, face edge, collar, folds, knees, and shoe tips to define the silhouette. Composition: full-body side view centered slightly right, one leg forward and one leg back, skirt and scarf dramatically swept by wind. Add a tall vertical strip of distressed pseudo-Japanese typography on the far left, like unreadable early-AI kanji printed in pale gray-beige blocks; include exactly 7 stacked text blocks, each separated by horizontal bars, with glitchy malformed characters rather than legible text. Add two thin vertical gray border lines near the left and right edges, plus a black ground shadow and horizontal base line under the feet. Style: high-contrast anime noir, screenprint poster, limited palette of red, black, cream, and muted gray, rough ink edges, graphic negative space, dramatic backlighting, slightly vintage printed texture. No realistic background, no extra characters, no readable modern text, no watermark.

Premium cinematic {argument name="角色" default="Western outlaw"} poster, vertical 2:3 composition, dark Western game character concept art poster style. A mysterious masked cowboy and a black horse standing at the desert border, full-body front view, wide-brimmed cowboy hat pulled low, patterned bandana covering the lower half of the face, dark long hair, black leather gloves, black Western jacket with multi-layered leather gear, bandolier, revolver holster, metal belt buckle, heavy boots, wearing a reddish-brown geometric patterned poncho on the shoulder with torn fluttering edges. The character's posture is calm and dangerous, one hand near the holster. A black horse enters the frame on the right, half-body, with a white forehead marking, clear rein details. The background is a Western desert in a storm, lightning, dark clouds, distant canyon walls, dead trees, dust, smoke, sparks, mud reflections, oppressive epic atmosphere. The left side of the image is vintage parchment negative space, the right side is a dark storm scene, strong light and dark division. Add large vertical English titles, wanted poster information, character profile, coordinates, map grids, thin wireframes, circular compass graphics, small red markers, signature stamps, and other premium poster layout elements. Style: black ink splatters, old paper texture, cinematic realism, dark Western, strong chiaroscuro, ultra-detailed leather and metal, dust, mud spots, scratches, smoke, sparks, rim lighting, premium collectible game poster, Red Dead Redemption vibe, artbook texture, 8K, high detail.

{argument name="resolution" default="3440*1440"} wallpaper, {argument name="subject" default="young japanese, girls group photo"}, {argument name="attire" default="school uniform"},

Using REFERENCE_0 as the source UI, redesign the Teams-like internal/external collaboration system into a cleaner, more modern product mockup while preserving the MPLUS Link brand, the dark navy left sidebar navigation, and the overall set of app functions. Create a polished 2-column by 3-row presentation showing exactly 6 redesigned screens: 1) a welcome/login landing screen with a large hero card, Japanese headline {argument name="headline text" default="社外パートナー連携を、招待制の専用ワークスペースで一元化"}, login tabs/buttons, and a realistic desktop-monitor hero image; 2) a project bridge chat screen with right-aligned chat cards, action chips, and a message composer; 3) a team list screen simplified into two clean team cards; 4) a materials/document list screen with upload controls, search/filter area, and file rows; 5) a shared collaboration room screen titled {argument name="room title" default="MPI 共同開発ルーム"} with a project overview card and compact member/status list; 6) a user management screen titled {argument name="admin title" default="ユーザー管理"} with user rows and an empty detail panel. Make the new mockups much more spacious than the reference: fewer table rows, larger margins, softer dividers, pale blue-gray backgrounds, white rounded cards, subtle shadows, and small blue/green accent buttons. Keep the Japanese enterprise SaaS feel and dummy-data nature, but remove the dense micromanaged tables and replace them with simplified, presentable mock UI suitable for a design-improvement comparison. Use brand name {argument name="brand name" default="MPLUS Link"} in the sidebar, keep the sidebar consistent on every screen, and avoid adding unrelated features, watermarks, or decorative clutter.

Create a realistic editorial print mockup photographed from above at a slight angle: a single white landscape-format card or magazine cover lying on a pale gray tabletop, with soft studio lighting, subtle paper texture, a thin shadow under the lower and right edges, and lots of clean negative space. The design is minimalist black-and-white with a small color accent. Place all typography flush left in the upper-left quadrant: a small italic serif kicker reading “{argument name="section label" default="Industry — Director Spotlight"}”, then a very large bold black modern sans-serif headline on three lines reading “{argument name="headline text" default="The Director-Led AI Studio Coming Out of London"}”, followed by a smaller gray-black sans-serif subheading on two lines reading “{argument name="subheading text" default="Future Video Media builds launch films, identity systems, and brand worlds in one week."}”. In the lower-left area beneath the copy, add exactly 1 black geometric film/play icon: a right-pointing triangular play symbol with three small square film perforations along its left edge. Under the icon add exactly 3 thin horizontal color stripes, stacked with small gaps: top stripe magenta fading into red and yellow, middle stripe deep blue, bottom stripe cyan/teal. Keep the card edges crisp, the background neutral, the overall style like a premium magazine feature cover or agency one-sheet, with no extra text, no people, no watermark, and no additional logos.

Using REFERENCE_1 as the base object and style reference, create a clean standalone illustration of the open Japanese flip phone, removing the surrounding infographic layout, captions, detail panels, borders, and all printed text. Replace the phone's screen content with the group photo from REFERENCE_0, fitted naturally inside the display as if it is being viewed in the phone gallery. Preserve the old silver-blue clamshell phone's angle, keypad, hinge, translucent worn plastic texture, small scratches, and blue star charm strap from REFERENCE_1. Keep the image in the same soft watercolor/anime concept-art style with a pale white background, faint blue shadows, and a few drifting translucent petals. The screen should show exactly 4 people from the group photo: blue-haired girl, black-haired boy, brown-haired girl, and brown-haired boy. Do not add UI text, labels, borders, page decorations, or extra objects beyond the phone, charm, subtle shadow, and petals. Optional customization: use {argument name="screen photo" default="the four-person summer group photo from REFERENCE_0"} on the phone display, with the phone model/color remaining {argument name="phone color" default="silver-blue translucent flip phone"}.

Create a dramatic monochrome ink illustration in the style of a gritty Japanese samurai manga: a lone {argument name="character type" default="ronin samurai"} stands in profile on a narrow wooden bridge spanning a rocky mountain gorge during heavy rain. The figure is a tall black silhouette with messy tied-up hair, a long tattered cloak whipping violently to the right, layered ragged robes, and two swords at the waist, one hand near the katana hilt. Use extreme black-and-white contrast, scratchy brushwork, splattered ink, and fine diagonal rain streaks across the entire scene. The bridge has wet planks, rough rail posts, and dark reflective puddles; below it, whitewater rapids churn between jagged cliffs. In the background, misty forested mountains fade into pale gray wash, with a small traditional Japanese temple or pavilion partly hidden among trees on the left cliff. Composition is vertical and cinematic, with the samurai slightly right of center, lots of pale empty rainy sky above, atmospheric fog, stormy motion, and no color except black, white, and gray. Mood: {argument name="mood" default="melancholic, lethal, and solitary"}.

Create a vertical 9:16 cinematic night portrait of {argument name="subject" default="a stylish young man"} standing on a high-rise balcony above a dense Tokyo city skyline, with Tokyo Tower glowing orange on the left and blurred skyscraper lights filling the background. The subject is framed from mid-thigh upward, leaning casually against a glass-and-metal railing with one arm draped over it, body angled slightly, wearing an elegant all-white outfit: oversized white blazer, loose open-collar white shirt, high-waisted white trousers, and a slim black belt, with a few subtle rings on the relaxed hand. The face is intentionally hidden by a centered dark square censor block, while dark medium-length hair remains visible around it. Use moody low-light photography, shallow depth of field, heavy bokeh, soft flash on the white suit, muted teal-and-orange color grading, slight film grain, chromatic aberration, VHS-like blur, and an atmospheric 1990s disposable-camera aesthetic. Make the background city lights dreamy and out of focus, the subject slightly soft but stylish, with a luxurious rooftop nightlife mood.

Create a stark black-and-white ink illustration of a {argument name="character type" default="wild samurai-like wanderer"} shown from the chest up against a blank white background, in a dramatic manga brush-pen style with rough crosshatching, heavy black shadows, splattered ink, and expressive scratchy linework. The figure wears a loose, dark, tattered kimono robe with a deep open collar exposing a long sinewy neck and upper chest, with ragged fabric edges and subtle worn floral markings on one side. The head is mostly covered by a large centered {argument name="face obscuration" default="flat medium-gray square censor block"}, but chaotic long {argument name="hair color" default="black"} hair bursts outward around it in many thin windblown strands, forming tangled arcs across the upper half of the image. Add exactly 4 black bird silhouettes around the figure: one large crow or raven flying in the upper left with wings spread, one small bird flying at lower left, one tiny feather-like black shape near the left middle, and one smaller crow silhouette at the right edge. Use high contrast, monochrome sumi-e energy, gritty seinen manga mood, no color except grayscale, no text, no frame, portrait composition, intense wind and motion, detailed ink texture on clothing and neck, minimal white negative space background.

Create a vertical cinematic night street photograph in rainy urban Japan, showing a {argument name="main subject" default="faceless young man in a white suit"} standing and walking slowly outside a brightly lit {argument name="store name" default="FamilyMart"} convenience store. He holds a transparent umbrella in one hand, with his other hand in his trouser pocket, wearing a white blazer, white slacks, white shirt, pale tie, and dark shoes; his face is intentionally blurred or obscured for anonymity. The scene is shot from sidewalk level with a slight wide-angle perspective, wet black pavement filling the foreground and reflecting neon signs, store lights, and street lamps. The storefront occupies the right side with glass doors, posters, trash bins, vending or ATM-like machines, and a green entrance mat; the large facade sign reads “FamilyMart.” On the upper left, include one hanging illuminated sign made of exactly 3 stacked panels labeled “FamilyMart,” “酒 たばこ,” and “銀行ATM.” Add a receding row of blurry illuminated shop signs and streetlights down the left side for depth. Use a moody neo-noir atmosphere, heavy rain, visible raindrops on the umbrella, glossy reflections, soft haze, analog film grain, slight chromatic aberration, muted teal-green and warm fluorescent tones, shallow depth of field, realistic photography, 35mm film look, subtle motion blur, high contrast, vertical 9:16 composition.

A cinematic single-panel anime illustration of a ruined mountain lodge interior at night, in a hand-painted concept-art style with dense ink linework, muted teal shadows and warm amber lamp light. Show {argument name="number of characters" default="four"} young travelers inside a weathered wooden cabin: one figure standing in the left foreground with their back turned, wearing an oversized winter parka with a fur-trimmed hood; one exhausted person slumped in a red armchair near the left-center; one seated person in an orange sweatshirt on the center sofa holding a mug, face partly obscured by shadow; and one quiet bob-haired girl standing on the right in a loose blue belted shirt-dress, socks, and shoes. The room is cluttered with old furniture, stacked crates, potted plants, blankets, cups, a low wooden coffee table, worn rugs, exposed beams, dangling wires, peeling walls, broken boards, and overgrown vines creeping through the structure. Place a large open wall or window at the back framing a cold blue view of {argument name="background landscape" default="snow-covered mountains under a starry night sky"}. Add two glowing paper lanterns on the left wall, casting orange light across the floor and characters while the rest of the cabin remains in cool blue-green darkness. Include a dark red bloodstain or ominous spill on the wooden floor in the foreground, creating a quiet mystery-thriller mood. Wide 16:9 composition, moody atmospheric perspective, intricate background details, painterly texture, dramatic contrast between warm interior light and icy mountain night, no text, no UI, no photorealism.

Using REFERENCE_0 as the base breakfast photo, transform it into a stylish social-media morning breakfast post. Keep the actual meal, plate, cup, and tabletop composition essentially the same, but remove the decorative sticker row at the bottom: exactly 6 sheep stickers and 1 pizza sticker must be gone. Add a warm, magazine-like food-photo finish with slightly cleaner lighting and sharper appetizing detail. Overlay casual white handwritten Japanese annotations directly on the photo, as if drawn with a marker pen, with thin arrows, dotted circles, hearts, underline strokes, and one cloud-shaped callout. Use exactly 10 visible annotation text groups: 1. Large top-left headline: {argument name="headline text" default="朝のトーストアート*"} 2. Small subtitle under it: {argument name="subtitle text" default="チーズと野菜のとろ〜りピザトースト"} 3. Left cloud callout about onion and melted cheese: 「シャキシャキ玉ねぎと とろけるチーズが 最高の相性！♡」 4. Small note near the green pepper with a dotted circle: 「ピーマンのほろ苦さがアクセントに◎」 5. Note pointing to the banana: 「バナナで自然な甘みをプラス♡」 6. Circular note near the egg: 「たんぱく質で一日を元気に！」 7. Note pointing to ketchup on the toast: 「ケチャップはたっぷりめがおいしいよね！」 8. Bottom-left motivational handwritten message with pink underline and heart: {argument name="motivational message" default="ちゃんと食べて、今日もごきげんにいこう。"} 9. Bottom-center note pointing to the toast crust: 「こんがりトーストにチーズがとろ〜り♡」 10. Bottom-right note pointing to the fruit bowl: 「フルーツとナッツでさっぱり＆カリッといいバランス♪」 Style constraints: make the handwriting look natural, playful, and hand-drawn; keep all text legible in white except the pink underline and heart accent at bottom left. Do not add any new food items, people, logos, watermarks, borders, or extra stickers.