GPT Image 2 prompts

Browse GPT Image 2 prompts for fashion portraits and graphic compositions. Compare how each prompt describes its subject and the way the image is arranged.

Explore prompts

10,039 prompts

Game Asset - Cyberpunk Gacha Game Dialogue Screenshot — output

Game Asset - Cyberpunk Gacha Game Dialogue Screenshot

iNFO DailyGPT Image 2
{"type":"gacha game urban exploration screenshot","style":"highly polished anime 3D render, cyberpunk action RPG UI, Zenless Zone Zero inspired, cinematic night lighting, detailed wet pavement reflections, neon signs, lantern glow, depth of field, 16:9 widescreen game capture","scene":{"setting":"narrow futuristic Asian city street at night with restaurants, storefronts, alleys, high-rise buildings, overhead string lights, red paper lanterns, glowing shop signs, arcade machine, street signs, café tables, and damp reflective asphalt","mood":"moody but vibrant, cozy neon nightlife, immersive open-world hub area","camera":"third-person dialogue view, eye-level camera, character centered slightly left, street receding into the distance"},"main character":{"name":"{argument name=\"character name\" default=\"INFO\"}","position":"foreground center-left","appearance":"young adult blonde woman with shoulder-length bob hair, slim build, wearing a loose white button-up dress shirt, black necktie with small yellow INFO lettering, dark skirt or trousers, one hand raised to her ear as if speaking through an earpiece","face":"covered by a flat tan rectangular censor block with no facial features","pose":"casual support-agent introduction pose, shoulders angled toward viewer"},"visible text":{"top left profile":"NEW! INFO Lv. 30","quest objectives count":2,"quest objectives":["Talk to Wise","Visit the video store"],"dialogue speaker":"???","dialogue lines count":3,"dialogue lines":["You must be the Proxy everyone’s talking about.","I’m with INFO.","If you ever need support, you know where to find us."],"distance marker":"26.0m","small sign":"Random Play","uid":"UID:13004538278"},"hud layout":{"top left":{"components count":4,"components":["black rounded main menu tile with white ME-like logo and red exclamation badge","yellow NEW! label above player name","player name and level block","two circular character/status icons"]},"top right":{"components count":6,"components":["Events icon labeled C","Store icon labeled F1","Tasks icon labeled F2","Agents icon labeled F3","Signal icon labeled F4","Menu icon labeled Tab"]},"right side":{"components count":4,"components":["circular minimap with compass N and quest icons","wireless scan button","objective navigation icon","26.0m distance label"]},"bottom left":{"components count":2,"components":["large transparent virtual joystick with W A S D labels","small camera shutter button labeled Q"]},"bottom right":{"components count":6,"components":["globe interaction button labeled R","large circular movement/skill button","ellipsis chat button labeled F","large swirl action button labeled Space","double-chevron dash button labeled Shift","small network bars beside UID"]},"bottom center":{"components count":1,"components":["black rounded dialogue box with white text, yellow highlighted character name, and double-chevron continue symbol"]}},"composition requirements":"Make it look like an authentic playable gacha-game screenshot rather than a poster. Keep the UI crisp and layered over the scene, with consistent icon styling, small keyboard shortcut tags, and game-like dialogue typography. Use {argument name=\"city color palette\" default=\"deep blue shadows, warm red lanterns, amber shop lights, and white neon highlights\"}. Dialogue should read exactly: {argument name=\"dialogue text\" default=\"You must be the Proxy everyone’s talking about. I’m with INFO. If you ever need support, you know where to find us.\"}. Render in {argument name=\"aspect ratio\" default=\"16:9\"}."}
Game Asset - Anime RPG Dialogue Screenshot — output

Game Asset - Anime RPG Dialogue Screenshot

iNFO DailyGPT Image 2
{
  "type": "gacha open-world RPG dialogue screenshot",
  "style": "high-end anime 3D game render similar to modern sci-fi urban titles, clean cel-shaded lighting, polished mobile game UI overlay, widescreen 16:9 gameplay capture",
  "scene": {
    "setting": "bright futuristic campus plaza with pale stone paving, broad stairs rising in the background, metal railings, elevated walkways, angular concrete-and-steel architecture, planters, rooftop greenery, blue sky, soft daylight, a few pink flowering trees in the distance",
    "camera": "third-person dialogue camera at chest height, centered NPC in foreground, wide environmental view, slight depth of field, cinematic but still clearly a gameplay screenshot",
    "mood": "welcoming, clean, sunny, futuristic academy-city atmosphere"
  },
  "main character": {
    "role": "NPC speaker",
    "name": "{argument name=\"character name\" default=\"Lotta\"}",
    "position": "center foreground",
    "appearance": "young anime woman with short blonde bob haircut, face intentionally obscured by a soft rectangular blur, fair skin, slim build",
    "outfit": "white button-up shirt with rolled sleeves, black necktie, black high-waisted pleated skirt, subtle school-uniform or office-uniform styling",
    "pose": "standing relaxed, one hand raised in a conversational open-palmed gesture, facing the viewer/player"
  },
  "secondary characters": {
    "count": 1,
    "description": "one small blonde NPC or player avatar visible at mid-right near a railing, walking away in the plaza"
  },
  "dialogue UI": {
    "count": 3,
    "elements": [
      "semi-transparent black label bar reading \"NPC Dialogue: Lotta\" above the dialogue text",
      "speaker name in yellow reading \"Lotta\"",
      "two-line white subtitle text reading \"Welcome! I've been waiting for you. It's truly an honor to finally extend our formal invitation to join the 'INFO' team.\""
    ],
    "dialogue text": "{argument name=\"dialogue text\" default=\"Welcome! I've been waiting for you. It's truly an honor to finally extend our formal invitation to join the 'INFO' team.\"}"
  },
  "HUD overlay": {
    "mobile game controls": "visible translucent virtual controls and action buttons over the 3D world",
    "top-left elements count": 5,
    "top-left elements": [
      "circular minimap",
      "small compass or navigation marker",
      "quest/document icon with red notification dot",
      "utility icon near the minimap",
      "top bar system icon cluster"
    ],
    "top-center elements count": 5,
    "top-center elements": [
      "small currency or resource number",
      "map/mission icon",
      "diamond-shaped menu icon",
      "camera or settings-style icon",
      "additional small menu glyph"
    ],
    "top-right elements count": 4,
    "top-right elements": [
      "red notification dot icon",
      "connection indicator showing 166ms",
      "character portrait of a pink-haired teammate",
      "character portrait of a blonde twin-bun teammate labeled Jiu"
    ],
    "right-side action buttons count": 7,
    "right-side action buttons": [
      "large running or dodge button",
      "large weapon attack button shaped like a blade",
      "crouch or slide movement button",
      "jump or movement skill button with wing-like icon",
      "small ability button above the attack cluster",
      "small circular interaction/skill button",
      "lower-right sprint or evade icon"
    ],
    "bottom elements count": 5,
    "bottom elements": [
      "chat bubble button at bottom-left",
      "faint build/version timestamp text",
      "center-bottom HP or stamina bar with numeric value",
      "small level indicator reading Lv. 90",
      "bottom-right user ID text"
    ]
  },
  "composition requirements": {
    "aspect ratio": "{argument name=\"aspect ratio\" default=\"16:9\"}",
    "render quality": "crisp anime game screenshot, detailed futuristic environment, readable UI, realistic sunlight and soft shadows",
    "avoid": "photorealism, medieval fantasy setting, painterly illustration, empty UI, close-up portrait crop, dark night lighting"
  }
}
Game Asset - Open-World Gacha Dialogue Screenshot — output

Game Asset - Open-World Gacha Dialogue Screenshot

iNFO DailyGPT Image 2
{
  "type": "open-world gacha game dialogue screenshot",
  "style": "high-fidelity anime character over realistic 3D urban environment, glossy mobile game UI, cinematic midday lighting, inspired by modern gacha RPGs such as Zenless Zone Zero, Wuthering Waves, Neverness to Everness and Arknights: Endfield",
  "scene": {
    "location": "wide city intersection with zebra crosswalk, glass office towers, storefronts, traffic lights, parked and moving cars, leafy green street trees, one distant pink flowering tree, clean asphalt road, bright blue sky with scattered clouds",
    "camera": "third-person dialogue view, waist-up to thigh-up character centered slightly left, 16:9 landscape mobile screenshot",
    "mood": "polished futuristic urban adventure, playable open-world RPG interface"
  },
  "main character": {
    "name in dialogue": "{argument name=\"character name\" default=\"Лотта\"}",
    "appearance": "adult anime woman with shoulder-length blonde hair parted near the center, slim build, confident posture, one hand raised behind her head, face intentionally covered by a soft beige rectangular blur with no facial features",
    "outfit": "white button-up dress shirt, loose black necktie, black high-waisted trousers, long black coat draped open over her shoulders, small ID badge clipped at the waist reading INFO, office-agent aesthetic",
    "hair color": "{argument name=\"hair color\" default=\"blonde\"}"
  },
  "layout": {
    "foreground": "large anime NPC standing on the crosswalk, overlapping the street scene and UI",
    "dialogue box": {
      "position": "bottom center",
      "style": "dark translucent rectangular box with thin gold accents and small decorative corner ornament",
      "speaker name": "{argument name=\"speaker label\" default=\"Лотта\"}",
      "dialogue text": "{argument name=\"dialogue text\" default=\"Привет! Ты ведь тоже ищешь возможности, верно? Тогда почему бы не заглянуть в iNFO? Я приглашаю тебя лично.\"}"
    },
    "hud components count": 11,
    "hud components": [
      "1 large circular minimap in the top-left corner with compass N, road grid, triangular player marker, car icon, menu icon with red notification badge, and small mascot icon",
      "1 bottom-left translucent virtual joystick with directional arrows",
      "1 small chat bubble button near the lower-left edge",
      "1 network and user ID readout at the bottom-left showing UID and latency",
      "1 top-right horizontal row of 7 small feature icons, several with red exclamation notification badges",
      "1 right-side vertical party list with exactly 3 circular anime character portraits labeled 1, 2, and 3",
      "1 center-right small translucent movement/look control circle",
      "3 large circular action buttons on the lower-right: one orbit/sprint-like icon, one running figure icon, and one two-person interaction icon",
      "1 smaller circular ability button attached near the right side of the dialogue box",
      "1 bottom-center status bar with level badge 1 and HP text 1250/1250",
      "1 small round red-gold skill icon at the bottom-right of the HP bar"
    ],
    "party portraits count": 3,
    "party portraits": [
      "slot 1: blue-haired anime character portrait with active highlight and small number 1",
      "slot 2: pink-haired anime character portrait with number 2",
      "slot 3: purple-haired anime character portrait with number 3"
    ]
  },
  "text details": {
    "store sign": "S COFFEE on a green storefront sign in the left background",
    "badge text": "INFO on the character's waist badge",
    "ui language": "Russian dialogue text, otherwise mostly icon-based interface"
  },
  "rendering instructions": "Create a convincing fake mobile game screenshot, not a poster. Keep the interface crisp and layered over the 3D city, with realistic perspective, screen-space transparency, sharp iconography, and the anime NPC composited naturally into the street environment."
}
Comic / Storyboard - Retro Bitrefill Diner Comic Ad — output

Comic / Storyboard - Retro Bitrefill Diner Comic Ad

RoddeGPT Image 2
{"type":"six-panel vintage comic advertisement","format":"vertical 2-column by 3-row comic page","style":"gritty retro American comic book illustration, heavy ink outlines, crosshatching, halftone texture, warm saturated colors, dramatic lighting, slightly weathered paper border, noir city diner atmosphere","characters":{"main character":"messy blond-haired young man in a red trench coat, red vest, white shirt and loose red tie, face intentionally covered by a flat brown rectangular censor block in every panel","server":"classic diner waiter in white short-sleeve uniform, black bow tie, and white paper hat, face intentionally covered by a flat brown rectangular censor block when visible"},"setting":"rainy nighttime city street and cozy neon-lit diner with chrome counter, red booths, tiled walls, condiment bottles, and glowing signs","layout":{"panel count":6,"arrangement":"two columns, three rows, thick black gutters","panels":[{"position":"top left","scene":"main character hunched over clutching his stomach outside a bright neon diner, wet street, car headlights, city skyline","visible text":["I'M STARVING!","DINER","GOOD FOOD!","OPEN 24 HOURS"]},{"position":"top right","scene":"close-up of the main character opening an empty wallet in panic outside the diner awning","visible text":["OH NO... I ONLY HAVE CRYPTO-- NO FIAT!"]},{"position":"middle left","scene":"heroic burst background as the main character holds up a green Bitrefill gift card with a five-dollar corner badge and crypto-themed icons, raising one finger confidently","visible text":["WAIT-- I HAVE MY BITREFILL CARD!","Bitrefill","Gift Card","$5","BITREFILL.COM"]},{"position":"middle right","scene":"inside diner at the counter, main character offers the green Bitrefill gift card to the waiter while pointing at a combo menu sign","visible text":["I'LL TAKE THE COMBO!","BEST VALUE!","COMBO","BURGER","FRIES","DRINK","$8 50","Bitrefill","Gift Card","$5"]},{"position":"bottom left","scene":"waiter serves a tray with a large burger, carton of fries, and striped soda cup to the main character at the counter","visible text":["PERFECT!","Good Food!"]},{"position":"bottom right","scene":"main character sits in a red booth happily eating a burger, fries and soda on the table, neon slogan in the window, bold ad banner at the bottom","visible text":["MMM-- THAT HIT THE SPOT!","Good Food! Good Life!","CRISIS AVERTED-- THANKS TO BITREFILL!"]}]},"product":{"brand":"{argument name=\"brand name\" default=\"Bitrefill\"}","item":"{argument name=\"product item\" default=\"Gift Card\"}","card color":"{argument name=\"card color\" default=\"green\"}","main slogan":"{argument name=\"main slogan\" default=\"CRISIS AVERTED-- THANKS TO BITREFILL!\"}"},"lettering":"hand-drawn bold uppercase comic lettering in cream speech bubbles, red emphasis words, thick black outlines, retro advertisement typography","composition notes":"make the food appetizing, the crypto-to-food narrative clear, include exactly 6 panels and preserve all visible speech bubbles and signs; faces remain hidden by simple opaque brown rectangles as if anonymized"}
Comic / Storyboard - Underwater Pink Petal Muse — output

Comic / Storyboard - Underwater Pink Petal Muse

우락갸GPT Image 2
Create a vertical fantasy anime illustration of a graceful underwater flower-woman standing barefoot on a pale sandy seabed, centered in the composition. She wears a simple satin-white slip dress with thin straps, softly reflective fabric, and a calf-length hem, while her body is enveloped by enormous flowing {argument name="flower petal color" default="pale blush pink"} petals resembling oversized ruffled sea anemone or oyster mushroom petals. The petals form a dramatic headdress and cloak around her, with many fine white vein lines, scalloped edges, translucent layering, and long trailing petal panels spreading across the floor like a gown train. Her face is intentionally hidden by a flat square {argument name="face cover color" default="dusty pink"} censor block placed over the head area. Set the scene in a dreamy underwater garden with soft blue-green water, sunbeams filtering down from above, floating bubbles, tiny glowing jellyfish, and blurred coral and anemone shapes in lavender, blue, and pastel pink in the background. Use delicate painterly rendering, soft gradients, luminous rim light, ethereal pastel colors, elegant fashion-illustration proportions, intricate petal texture, serene magical atmosphere, and high detail.
Profile / Avatar - Dreamy Underwater Pink-Haired Portrait — output

Profile / Avatar - Dreamy Underwater Pink-Haired Portrait

우락갸GPT Image 2
Create a dreamy underwater fantasy portrait of a young woman from mid-thigh upward, standing still against a deep teal ocean background with tiny suspended particles and soft haze. She has very long, flowing, wavy {argument name="hair color" default="pastel pink"} hair drifting weightlessly around her shoulders, with a coral-like hair ornament tucked on the right side of her head. She wears a delicate semi-sheer {argument name="dress color" default="ivory white"} sleeveless slip dress with thin straps, a softly gathered waist, and gauzy fabric that looks ethereal underwater. Her skin is pale and softly lit, her posture is relaxed with both arms hanging down, and the composition is vertical, centered, and painterly-realistic with a muted cinematic glow. Add an opaque square censor block over the entire face, colored {argument name="censor block color" default="dusty mauve beige"}, obscuring all facial features while leaving the hair and body visible. Use soft focus, low contrast, subtle bloom, underwater diffusion, fine art fantasy photography mood, no text, no extra objects.
Infographic / Edu Visual - Magical Creature Bestiary Plates — output

Infographic / Edu Visual - Magical Creature Bestiary Plates

波特er队长GPT Image 2
{"type":"high-end editorial magical creature bestiary infographic page set","overall_style":"antique wizarding-world natural history archive, aged parchment paper, burned and torn edges, sepia ink, engraved field-guide illustration mixed with realistic painterly creature rendering, medieval manuscript ornaments, compass roses, red wax seals, rune-like marginalia, hand-written researcher notes, museum dossier layout, premium fantasy encyclopedia design, highly detailed, cinematic but flat document presentation","format":"three illustrated bestiary plates arranged as one composite image: one large horizontal plate on top and two smaller vertical plates below","parameters":{"creature 1 name":"{argument name=\"creature 1 name\" default=\"Hippogriff\"}","creature 2 name":"{argument name=\"creature 2 name\" default=\"Niffler\"}","creature 3 name":"{argument name=\"creature 3 name\" default=\"Chinese Fireball Dragon\"}","archive series title":"{argument name=\"archive series title\" default=\"Magical Creatures Archive\"}","paper tone":"{argument name=\"paper tone\" default=\"weathered tan parchment\"}"},"layout":{"total_plates":3,"plate_1":{"position":"top full width","title":"{argument name=\"creature 1 name\" default=\"Hippogriff\"}","main_subject":"large majestic hippogriff with eagle head, hooked beak, sharp golden talons, feathered chest, horse-like hindquarters, long tail, and one enormous wing spread across the upper right of the page","classification":"XXXXX","visible_sections_count":9,"visible_sections":["opening description paragraph","ANATOMY OVERVIEW","MAGICAL ADAPTATIONS","BEHAVIOR TRAITS","DANGER LEVEL","HABITAT","DIET","KNOWN THREATS","HANDLING GUIDELINES"],"danger_level":"5 red diamond scale, moderate to high if provoked","habitat_visual":"small circular mountain habitat map with cliffs and pine forest","handling_guidelines_icons_count":5,"handling_guidelines_labels":["Approach Slowly","Bow Your Head","Maintain Respect","Do Not Make Eye Contact","Never Use Force"],"notes":"include a pinned handwritten field note and a Ministry Remark note sealed with a red stamp"},"plate_2":{"position":"bottom left","title":"{argument name=\"creature 2 name\" default=\"Niffler\"}","main_subject":"cute black furry mole-like creature with glossy eyes and duckbill snout sitting amid scattered gold coins, chains, jewels, treasure boxes, and a small pouch","visible_sections_count":8,"visible_sections":["description paragraph","ANATOMY","CLASSIFICATION / ORIGIN / MAGICAL RESISTANCE / DANGER LEVEL","BEHAVIOR","HABITAT","DIET","HANDLING NOTES","WARNING"],"danger_level":"3 red diamonds out of 5","origin":"Britain","magical_resistance":"Low","habitat_visual":"small underground tunnel and bank-vault style habitat map","diet_items_count":4,"diet_items":["insects","worms","small creatures","anything edible"],"handling_notes_count":4,"handling_notes":["Keep valuables secured","Approach slowly","Use a charm bag if necessary","Never stand near a Niffler"],"warning":"May cause property damage; not suitable for inexperienced handlers"},"plate_3":{"position":"bottom right","title":"{argument name=\"creature 3 name\" default=\"Chinese Fireball Dragon\"}","subtitle":"中国火球龙","main_subject":"long serpentine red and gold Chinese dragon coiled in an S shape, horned head, whiskers, spines along the back, glowing fireballs orbiting around it, smoky ember effects","visible_sections_count":8,"visible_sections":["ANATOMY","DANGER LEVEL","HABITAT","MAGICAL ABILITIES","DIET","BEHAVIOR","RESEARCHER'S NOTE","classification label"],"danger_level":"5 red diamonds, EXTREME","habitat":"mountain ranges in China, prefers high altitudes and remote caves","magical_abilities_count":4,"magical_abilities":["Fireball Projection","High Temperature Tolerance","Rapid Flight","Enhanced Senses"],"behavior_traits_count":4,"behavior_traits":["Highly territorial","Aggressive when provoked","Protects its nest fiercely","Respected and feared in magical cultures"],"researcher_note":"Extremely dangerous. Only experienced magizoologists should attempt observation."}},"typography":"large black serif display titles, small uppercase section headers, fine ruled lines, red diamond danger markers, tiny pseudo-scholarly labels and catalog numbers, old ink annotations","composition_requirements":"make the pages feel authentic and information-dense, with exact readable English headings where possible; include anatomical sketches, small diagrams, maps, icons, and marginal notes without modern UI elements","lighting_and_texture":"soft overhead scan lighting, parchment stains, creases, ink smudges, foxing, scratches, taped corners, archival wear","quality":"ultra-detailed fantasy infographic, sharp linework, realistic creature anatomy, coherent editorial layout, no photorealistic humans"}
Comic / Storyboard - Labanotation to 4-Beat Pose Sheet — output

Comic / Storyboard - Labanotation to 4-Beat Pose Sheet

WTRGPT Image 2
Using the provided reference image as a Labanotation-style dance score, interpret the 4-beat movement phrase and convert it into a pose-control character sheet. Create exactly 4 full-body panels arranged in a clean 2×2 grid on a plain white studio background, showing the same character performing the pose for each beat. Add bold Japanese beat labels in the upper-left of each panel: "1拍目", "2拍目", "3拍目", and "4拍目". Use a photorealistic young female model wearing a {argument name="outfit" default="white school-uniform blazer, black bow tie, gray plaid skirt, white socks, and black Mary Jane shoes"}, with {argument name="hairstyle" default="black twin tails"}. Translate the notation into these four poses: beat 1: standing on one leg with one arm raised straight upward and the opposite leg bent back; beat 2: one arm raised upward, body angled slightly, one knee lifted and crossed in front; beat 3: standing front-facing with both arms extended horizontally to the sides; beat 4: standing front-facing with both hands held together near the chest. Keep the character consistent across all four panels, use soft even lighting, minimal shadows, and a simple catalog-like dance reference layout.
Product Marketing - Japanese Veranda Summer Portrait — output

Product Marketing - Japanese Veranda Summer Portrait

俺の娘たち / プロンプトの人GPT Image 2
Create a photorealistic vertical editorial portrait of an adult woman relaxing on the wooden veranda of a traditional Japanese house, seated on a muted blue floor cushion beside tatami mats and open sliding shoji doors. She wears {argument name="outfit" default="white resort knit beachwear with a ribbed triangle top and matching bottoms"} under a sheer oversized white linen shirt that slips loosely off one shoulder, with delicate dewy skin and faint moisture catching the sunlight. Her dark hair is tied back casually, her face is intentionally anonymized by a simple square blur, and she holds {argument name="drink" default="a green canned drink with Japanese-style labeling"} near her mouth in a quiet summer moment. Behind her, translucent noren curtains hang in the doorway, colored white with soft indigo-blue wave and bird motifs, moving gently in the breeze; outside is a lush Japanese garden with moss, stepping stones, leafy greenery, hydrangeas, bamboo, and dappled afternoon sun. Add a small hanging glass wind chime near the eaves, warm natural wood textures, soft backlighting, realistic skin highlights, shallow depth of field, cinematic summer atmosphere, high-end lifestyle photography, 9:16 composition, natural colors, serene and intimate but tasteful mood.
YouTube Thumbnail - Melancholic Guitar Girl and Cat — output

YouTube Thumbnail - Melancholic Guitar Girl and Cat

WTRGPT Image 2
Create a wide minimalist anime illustration with abundant empty negative space on the left and the main subject placed on the right. Show a quiet teenage girl in full body side profile facing left, with {argument name="hair color" default="dark navy-black"} short bob hair, blunt bangs, pale skin, and a calm melancholic expression. She wears an oversized dark hooded jacket over a dark plaid school-uniform style outfit, a short pleated plaid skirt, black fishnet tights with a small black garter bow on one thigh, and chunky black platform shoes. A tall dark guitar gig bag is strapped vertically on her back, extending above her head. In one relaxed hand she holds a smartphone. At her feet sits exactly 1 black cat in silhouette, looking up at her. Use a soft off-white background with pale blue-gray watercolor cloud washes behind the figures, delicate ink linework, muted blue-green shadows, subtle cel shading, airy composition, quiet indie music mood, high-detail anime key visual style, no text, no border.
Comic / Storyboard - Japanese MV Key Visual Sheet — output

Comic / Storyboard - Japanese MV Key Visual Sheet

MOSAiC🎧AI邦RockGPT Image 2
{"type":"Japanese music video key visual sheet and character design board","title_text":"{argument name=\"main Japanese title\" default=\"となりあわせ\"}","subtitle_text":"{argument name=\"subtitle text\" default=\"— となりに、君がいる日常 —\"}","overall_style":"warm cinematic indie J-pop MV planning sheet on a clean white editorial canvas, beige grid lines, small serif English headings, Japanese copy, soft natural window light, nostalgic apartment atmosphere, muted cream and terracotta palette, realistic photography mixed with production-design reference-board layout","main_key_visual":{"position":"upper left, largest panel","description":"a young Japanese man with short dark slightly wavy hair sits on a sofa in a cozy sunlit room playing an acoustic guitar, wearing a rust-orange oversized knit cardigan over a white T-shirt and loose ivory pants; background includes curtains, houseplants, books, desk lamp, cushions and warm bokeh; intimate shallow depth of field, calm melancholy mood","visible_text":{"large_title":"{argument name=\"main Japanese title\" default=\"となりあわせ\"}","subtitle":"{argument name=\"subtitle text\" default=\"— となりに、君がいる日常 —\"}","body_copy":"small vertical stack of Japanese poetic copy in white"}},"character":{"section_title":"CHARACTER - 全身図(衣装:MVメイン)","position":"upper right","name":"{argument name=\"character name\" default=\"unnamed young Japanese male singer-songwriter\"}","appearance":"slim young adult male, short black hair, gentle indie musician impression","main_outfit":"{argument name=\"outfit description\" default=\"rust-orange cardigan, white crewneck T-shirt, long checked shirt layer, loose ivory trousers, brown-and-white sneakers\"}","turnaround_count":4,"turnaround_views":["FRONT","SIDE","BACK","3/4"],"turnaround_description":"four full-body studio cutouts on white showing the same outfit from front, side, back and three-quarter angles"},"expression_variations":{"section_title":"表情バリエーション(感情の変化)","position":"middle right below character turnaround","count":10,"labels":["微笑み","無表情","悲しみ","驚き","優しさ","考える","照れ","寂しさ","決意","目を閉じる"],"description":"ten close-up head-and-shoulders crops of the character, arranged in two rows of five, each labeled in Japanese and showing subtle emotional changes while keeping soft natural skin tones and dark hair"},"text_blocks":{"concept":{"heading":"CONCEPT","position":"lower left under main visual","description":"short Japanese paragraph explaining a quiet everyday love-story concept about being next to someone, loneliness, warmth and small happiness"},"mood_and_tone":{"heading":"MOOD & TONE","position":"lower center-left under main visual","description":"Japanese hashtag-like mood words in terracotta ink emphasizing warmth, daily life, tenderness, nostalgia, natural light and emotional distance"}},"sub_images_mood_cuts":{"section_title":"SUB IMAGE / MOOD CUTS","position":"horizontal strip across the lower middle","count":7,"items":["close side-profile crop of the guitarist in orange cardigan","sunlit table with glass vase and small plant near a window","cozy bedroom or living corner with warm lamp light and curtains","orange patterned mug beside notebook or papers on a desk","soft close-up of the character looking upward near a window","overhead desk still life with notebook, pen, sketches and guitar edge","profile shot of the character facing bright window light with greenery"]},"bottom_sections":{"world_view":{"heading":"WORLD VIEW","position":"bottom left","thumbnail_count":2,"description":"Japanese world-building text plus two small warm interior thumbnails showing morning window light and nighttime amber room light"},"color_palette":{"heading":"COLOR PALETTE","position":"bottom center-left","swatch_count":6,"swatches":["terracotta orange","soft peach beige","warm ivory","muted sage green","dark coffee brown","pale gray-white"],"description":"six rectangular color swatches with small hex-code labels beneath"},"symbol_and_motif":{"heading":"SYMBOL & MOTIF","position":"bottom center","count":5,"items":["acoustic guitar","orange patterned mug","small houseplant","sheer curtains with sunlight","notebook and pencil"]},"keywords":{"heading":"KEYWORDS","position":"bottom center-right","count":6,"items":["日常","となりあわせ","やさしさ","記憶","つながり","光と影"],"description":"vertical Japanese keyword list with a tiny botanical line drawing in terracotta"},"light_and_atmosphere":{"heading":"LIGHT & ATMOSPHERE","position":"bottom right","count":3,"items":["soft morning window light through plants","golden curtain glow in a dim room","nighttime candle or lamp-lit interior"],"description":"three small cinematic lighting-reference thumbnails with explanatory Japanese copy"}},"layout":"wide 16:9 presentation board, white margins, thin divider lines, structured editorial grid, main key visual dominates left half, character sheet dominates right half, reference thumbnails and palette run along the bottom","rendering_notes":"photorealistic, cinematic stills, soft grain, warm highlights, shallow depth of field, tasteful Japanese typography, do not make it look like a comic; faces may be softly anonymized or minimally detailed if needed while preserving hairstyle and pose"}
Social Media Post - NBA Courtside Broadcast Screenshot — output

Social Media Post - NBA Courtside Broadcast Screenshot

DanFeiGPT Image 2
{"type":"realistic NBA broadcast screenshot photo","scene":"courtside crowd at a packed professional basketball arena during a live game, wide 16:9 TV broadcast frame, shallow depth of field, crisp sports photography, bright arena lighting, blue LED ribbon boards and blurred spectators in the background, hardwood court visible on the right with distant players in white and blue uniforms","main_subject":{"description":"a glamorous female courtside fan seated in the front row, centered slightly right, leaning forward toward the camera with confident celebrity-like presence","face":"intentionally obscured by one large opaque pixelated blur rectangle over the entire face for anonymized viral-template style","hair":"{argument name=\"hair color and style\" default=\"long glossy black wavy hair parted near the middle\"}","outfit":"{argument name=\"main subject outfit\" default=\"off-shoulder white top with thin straps, delicate layered necklace, bare shoulders\"}","pose":"seated close to camera, shoulders angled, relaxed courtside posture"},"surrounding_people":{"count":7,"description":"seven visible nearby spectators also anonymized with soft rectangular blur blocks over their faces","people":["left edge man in black Clippers hoodie, cropped face partly visible","rear-left standing fan in white shirt with large face blur","small rear-left fan wearing blue LA cap with vertical face blur","middle-left spectator in white shirt with small face blur","right-middle spectator behind main subject with small face blur","right-front man in black LA Clippers Basketball shirt and dark cap with face blur","far-right partial spectator with small face blur"]},"broadcast_graphics":{"count":4,"components":[{"position":"top-left","description":"NBA/TNT scoreboard overlay with two teams and game clock","visible_text":["GSW","68","LAC","80","3:21","3RD","24","TNT"]},{"position":"top-right","description":"network watermark","visible_text":["SPECTRUM","SPORTSNET","LIVE"]},{"position":"bottom-left to bottom-center","description":"social media lower-third ticker bar with Twitter icon, handle, caption, fire and raised-hands emojis, and hashtag","visible_text":["@BballFanatic24","She’s locked in! Best courtside energy 🔥🙌 #ClipperNation"]},{"position":"bottom-right","description":"camera recording HUD label with corner brackets and red recording dot","visible_text":["CAM 02","REC"]}]},"sports_context":{"teams":"{argument name=\"teams on scoreboard\" default=\"GSW vs LAC\"}","score":"{argument name=\"score and clock\" default=\"GSW 68, LAC 80, 3:21 remaining in 3rd quarter, 24 shot clock\"}","arena_details":"Los Angeles Clippers courtside atmosphere, black Clippers hoodies, towels stacked near seats, packed crowd, broadcast camera angle as if from sideline media row"},"style":"hyper-realistic candid TV sports broadcast still, authentic NBA arena atmosphere, high-resolution photojournalism, realistic skin tones and fabric textures, slight motion blur in distant background, dramatic courtside perspective, viral social-media sports screenshot aesthetic","composition":"main subject dominates foreground center, scoreboard in upper left, watermark in upper right, lower-third caption spans the bottom, REC overlay at lower right, court and players recede into background on the right","negative_prompt":"no illustration, no cartoon, no fantasy uniforms, no extra limbs, no distorted hands, no readable brand errors beyond broadcast overlays, no unblurred faces for background spectators"}
Social Media Post - NBA Courtside Fan Cam Broadcast — output

Social Media Post - NBA Courtside Fan Cam Broadcast

DanFeiGPT Image 2
{"type":"ultra-realistic NBA on TNT broadcast screenshot photo","scene":"courtside fan cam during a live professional basketball game, shot from a TV broadcast camera with shallow telephoto compression, arena crowd filling the background, realistic stadium lighting, authentic sports broadcast color grading, 16:9 horizontal frame","main_subject":{"position":"center foreground, leaning forward over courtside seats","description":"{argument name=\"central fan description\" default=\"glamorous adult woman with long dark wavy hair wearing a white off-shoulder cropped top, seated courtside and looking toward the camera\"}","face":"covered by a soft rectangular privacy blur, matching the anonymized crowd style"},"crowd":{"description":"dense courtside audience behind and around the subject, fans standing, clapping, holding drinks, wearing mixed basketball apparel","visible_team_apparel_count":14,"team_apparel_breakdown":["4 Lakers-themed shirts or jerseys in yellow/purple","6 New York Knicks-themed shirts or jerseys in blue/orange/white","4 Brooklyn Nets-themed shirts or jerseys in black/white"],"privacy_blurs":{"count":24,"style":"soft rectangular face-obscuring blur blocks in skin-tone colors placed over individual faces throughout the crowd"}},"broadcast_graphics":{"count":5,"components":[{"name":"top-left scoreboard","content":"NBA logo vertical strip, 4TH QTR, game clock {argument name=\"game clock\" default=\"8:45\"}, teams {argument name=\"scoreboard teams and score\" default=\"LAL 105, BKN 101\"}, red shot-clock box reading SHOT 22"},{"name":"top-right network watermark","content":"semi-transparent circular TNT logo"},{"name":"center-lower segment label","content":"NBA ON TNT | COURTSIDE FAN CAM"},{"name":"bottom headline ticker","content":"{argument name=\"headline text\" default=\"LEBRON JAMES WITH A MASSIVE DUNK... KEVIN DURANT RESPONDS...\"}"},{"name":"bottom social ticker","content":"Twitter-style icon followed by @SportsFanatic: {argument name=\"social ticker quote\" default=\"Check out the courtside energy! #NBAPlayoffs\"}"}]},"composition":"make it look like a paused live TV broadcast screenshot, with the subject centered, crowd cropped naturally at the edges, courtside seats visible at the bottom, scoreboard occupying the upper-left corner, watermark in the upper-right corner, and two stacked lower-third bars across the bottom","style":"photorealistic, high-detail broadcast still, realistic arena depth, slight compression artifacts, clean sharp typography, convincing TNT/NBA-style graphics, no illustration or cartoon look"}
YouTube Thumbnail - Warm Firelit Portrait Transformation — output

YouTube Thumbnail - Warm Firelit Portrait Transformation

hxhxhx0916GPT Image 2
Using the provided reference image as the identity reference, create a warm low-angle cinematic photo of the same man, keeping his face and natural light stubble consistent with the reference. Change the scene so he is reclining in a leather armchair at night, wearing a thin, slightly wrinkled {argument name="shirt color" default="light beige"} button-up shirt instead of the suit. Pose him relaxed and leaning back, with one hand holding a short glass of {argument name="drink" default="whiskey"}. Add a large out-of-focus {argument name="foreground light source" default="orange fire flame"} on the right side of the frame, casting flickering amber light and shadows across his face and shirt. Set the background in a dark, softly blurred intimate interior with bookshelves and a warm lamp, using amber and brown tones for a melancholic, nostalgic evening mood. Make it look like a high-ISO modern digital camera or smartphone photo with visible grain, low contrast, shallow depth of field, and subtle chromatic aberration.
Comic / Storyboard - Celestial Sky God Engraving — output

Comic / Storyboard - Celestial Sky God Engraving

defyongGPT Image 2
Create a vertical mythic fantasy engraving in the style of a 19th-century Gustave Doré woodcut, rendered in dense black-and-cream linework with fine cross-hatching, swirling etched strokes, and aged parchment texture. The scene shows a tiny lone traveler, {argument name="traveler description" default="a cloaked wanderer holding a walking staff"}, standing near the bottom center of a vast barren desert plain, facing an enormous celestial apparition in the sky. Above the horizon, clouds and wind spiral into the monumental bust of {argument name="deity identity" default="an ancient bearded sky god"}, with stern glowing eyes, flowing hair and beard made of cloud and storm lines, an ornate crown, circular halo rings behind the head, and one giant hand emerging from the clouds in a calm blessing or summoning gesture. The landscape is wide and desolate, with rippled sand, scattered rocks, sparse shrubs, distant low mountains, and a bright open horizon that emphasizes the scale difference between the human and the divine figure. Use a cinematic low viewpoint, epic vertical composition, surreal visual storytelling, sacred atmosphere, storm-cloud vortexes, intricate engraved texture, monochrome ink with subtle blue-gray shading, high detail, dramatic contrast, no modern objects, no readable text except an optional tiny artist signature near the bottom reading {argument name="signature text" default="x.com/defyong"}.
Infographic / Edu Visual - Tomato Cauliflower Anime Comic — output

Infographic / Edu Visual - Tomato Cauliflower Anime Comic

Adam也叫吉米GPT Image 2
Using REFERENCE_0 as the cauliflower character source and REFERENCE_1 as the tomato character source, transform the ordinary market produce into a cute children's anime story comic. Anthropomorphize one red tomato and one white cauliflower as expressive chibi friends with faces, arms, legs, and lively emotions, while keeping their recognizable vegetable shapes and colors. Create a warm 3x3 storyboard grid with exactly 9 numbered panels, rounded borders, cozy market/kitchen lighting, speech bubbles and narration boxes, in a polished picture-book anime style. The story should show them meeting in the vegetable market, becoming friends, washing and preparing ingredients, cooking together in a pan, creating a colorful tomato-cauliflower dish, tasting it, and ending with a dream of becoming tiny super-chef heroes. Include the following 9 panel texts in Chinese as visible captions/dialogue: {argument name="nine panel Chinese story text" default="1 在菜市场里,西红柿和花菜相遇了。 你好吗! 嗨~ | 2 西红柿好奇地问:‘你为什么长得像小树呀?’ 因为我有花椰朵朵呀~ | 3 ‘我们一起做菜吧!’ 西红柿提议。 好呀好呀! | 4 他们洗干净,切一切,准备大显身手! | 5 下锅啦~滋滋滋——香味四溢! | 6 色香味俱全的‘西红柿炒花菜’出锅啦! | 7 尝一口,嗯~真好吃! | 8 从此,他们成了最好的朋友! | 9 他们还有一个小小的梦想:成为超级美味料理侠!"}. Make the tone cheerful, educational, and suitable for young children, with bright colors, sparkles, hearts, and soft cinematic lighting. Use the reference images only as ingredient inspiration; do not keep the original photo background except where reimagined as a cozy illustrated vegetable market.
Infographic / Edu Visual - Diaochan 3D Break-Frame Archive Poster — output

Infographic / Edu Visual - Diaochan 3D Break-Frame Archive Poster

Larus CanusGPT Image 2
{"type":"3D Chinese historical beauty break-frame character archive poster","subject":{"character name":"{argument name=\"character name\" default=\"貂蝉\"}","era":"Eastern Han, Three Kingdoms romance aesthetic","appearance":"peerlessly beautiful noblewoman, graceful dancer posture, ornate high updo with gold hairpins, pink blossoms, dangling pearl ornaments, delicate makeup, serene enchanting expression","costume":"layered Han-style silk robes in ivory, blush pink, rose red and deep burgundy, translucent flowing sleeves and ribbons, embroidered floral hems, gold chains, waist sash, jade-and-gold accessories","held object":"{argument name=\"held object\" default=\"a carved ivory crescent moon hand mirror with tassels and gold ring charms\"}","pose":"full-body statue-like figure stepping forward from an antique golden picture frame, right hand raised elegantly, left arm extended toward the viewer holding the object so it breaks out of the frame"},"composition":{"format":"vertical poster, cinematic product-archive layout, single central full-body figure surrounded by documentary cards","central effect":"shattered glass and cracked golden frame behind the character, translucent shards floating outward, parts of the robe, ribbons and held object protrude past the frame for strong 3D depth","background":"aged parchment texture with misty ancient Chinese palace silhouettes, full moon, faint calligraphy columns, sepia ink-wash mountains and architecture","color palette":"warm parchment beige, antique gold, smoky gray, rose pink, burgundy, ivory, soft moonlight","lighting":"dramatic soft frontal light, rim highlights on glass shards and gold frame, glowing particles, high-detail realistic 3D render mixed with Chinese ink-painting collage"},"typography":{"large title":"{argument name=\"large title text\" default=\"貂蝉\"}","subtitle":"{argument name=\"subtitle text\" default=\"东汉末·闭月\"}","left vertical seal":"small red Chinese seal below subtitle","bottom plaque":"{argument name=\"bottom plaque text\" default=\"一拜明月,乱世生谋\"}"},"layout":{"left side name plaques":{"count":3,"labels":["司徒府歌姬","貂蝉拜月","连环计"]},"story cards":{"count":8,"style":"small parchment-and-gold bordered cards with circular number badges, miniature monochrome historical illustrations, short Chinese captions, connected by thin glowing geometric lines","cards":[{"number":"01","position":"upper left","title":"乱世将倾 京师风云","image":"ancient capital and palace roofs in mist"},{"number":"02","position":"left upper middle","title":"司徒府中 歌舞惊才","image":"young woman performing before an elder"},{"number":"03","position":"left middle lower","title":"月下焚香 貂蝉拜月","image":"woman worshipping beneath a moonlit pavilion"},{"number":"04","position":"lower left","title":"连环计起 以柔制暴","image":"strategic meeting with officials"},{"number":"05","position":"upper right","title":"董卓权倾 朝堂阴影","image":"dark court scene with powerful warlord"},{"number":"06","position":"right upper middle","title":"吕布动心 父子反目","image":"mounted warrior and dramatic confrontation"},{"number":"07","position":"right middle lower","title":"凤仪亭事 情势决裂","image":"garden pavilion confrontation"},{"number":"08","position":"lower right","title":"董卓伏诛 传说永存","image":"aftermath scene with figures in historical robes"}]},"bottom timeline":{"count":5,"labels":["乱世","拜月","连环","凤仪","伏诛"],"style":"thin gold horizontal line with circular milestones"}},"rendering instructions":"ultra-detailed 3D character, realistic silk translucency, intricate embroidery, ornate carved stone pedestal, antique gold UI borders, documentary archive infographic, dynamic flowing fabric, depth of field, high resolution, no modern objects, no English text in the poster, preserve readable Chinese calligraphy labels"}
Profile / Avatar - Cozy Bedroom Selfie Portrait — output

Profile / Avatar - Cozy Bedroom Selfie Portrait

ギンGPT Image 2
Create an ultra-realistic vertical smartphone selfie photo of a cute young woman in a cozy softly lit bedroom, shot from a high slightly downward angle with her arms extended toward the camera as if taking the picture herself. She has {argument name="hair color" default="long glossy black hair"} styled in loose waves with a black ribbon bow hair clip on the left side, pale skin, and a delicate feminine look; her face is intentionally covered by a large centered pixelated square privacy blur in muted brown-gray tones. She wears a {argument name="outfit color" default="white"} fitted camisole with lace trim under a fluffy fuzzy cardigan, plus a small pearl pendant necklace. The composition is close and intimate, showing her upper body and arms prominently in the foreground. The background is a tidy warm-toned bedroom with an unmade cream bed on the left, a small shelf or dresser holding cosmetics and a glowing candle or lamp, two small pastel wall prints, tall beige wardrobe doors on the right, and a cream handbag hanging from a hook. Use natural indoor lighting, soft shadows, realistic skin and fabric textures, casual Japanese social-media selfie aesthetic, shallow depth of field, 9:16 portrait framing, high-resolution phone camera realism.
Profile / Avatar - Realistic Mirror Selfie in Apartment Entryway — output

Profile / Avatar - Realistic Mirror Selfie in Apartment Entryway

ギンGPT Image 2
Create an ultra-realistic vertical smartphone mirror selfie of a young woman in a small apartment entryway. She has long, slightly messy wavy {argument name="hair color" default="dark brown to black"} hair falling past her shoulders, with her face intentionally hidden by a solid square privacy block in the center. She is holding a {argument name="phone color" default="black"} smartphone in her right hand at face height, taking the photo in a mirror, while her left hand is tucked casually into her jacket pocket. She wears an oversized {argument name="jacket color" default="black"} zip-up hoodie left open over a plain {argument name="shirt color" default="white"} T-shirt, with loose dark pants, giving a relaxed casual streetwear look. The setting is a narrow Japanese-style entrance hallway with beige walls, a gray door and metal latch behind her, pale tiled flooring, a wooden mirror frame on the right edge, a light switch on the wall, and exactly 3 visible pairs of shoes: one white sneaker pair on the left floor, one glossy black shoe pair near the right wall, and one dark shoe pair partly cropped at the bottom right. Use warm indoor lighting, realistic phone-camera perspective, natural imperfect composition, slight grain, soft shadows, and candid social-media realism, as if photographed with a handheld phone rather than professionally staged.
Profile / Avatar - Cafe Selfie With Privacy Block — output

Profile / Avatar - Cafe Selfie With Privacy Block

ギンGPT Image 2
Create an ultra-realistic vertical smartphone selfie of a cute young Japanese woman sitting in a warm modern cafe. She has {argument name="hair color" default="light chestnut brown"} medium-length softly layered wavy hair with airy bangs, glossy natural texture, and loose curls falling around her shoulders. Her face is intentionally covered by a centered opaque rectangular privacy block in a warm brown skin-tone color, hiding all facial features. She wears an oversized {argument name="sweater color" default="cream white"} ribbed knit cardigan with large buttons and a deep V neckline, one long sleeve pulled up near her mouth in a shy pose, with a thin black shoulder bag strap visible on one shoulder. Frame it as a close-up chest-up selfie from a slightly high angle, natural casual posture, realistic body proportions, soft indoor lighting, shallow depth of field, high-end phone camera realism. The background is a stylish cafe interior with wooden tables, glass partitions, round glowing pendant lamps, blurred storefront lights, and a leafy indoor plant on the right. Use warm beige tones, cozy atmosphere, candid social-media snapshot aesthetic, photorealistic skin and fabric detail, no illustration style, no text, vertical 9:16 composition.
Profile / Avatar - Luxury Boutique Two-Woman Selfie — output

Profile / Avatar - Luxury Boutique Two-Woman Selfie

ギンGPT Image 2
Create an ultra-realistic vertical smartphone selfie photo of two fashionable young women standing close together in an elegant, brightly lit luxury boutique or salon interior. The camera is held slightly above chest height at arm’s length, creating a casual wide-angle selfie perspective with both women cropped from the head to mid-thigh. Both faces are intentionally covered by large opaque rectangular privacy blocks with soft brown-to-rose gradient tones, centered over the face area. The woman on the left has {argument name="left woman hair color" default="dark brown"} long straight hair with a small top section tied back, wearing a cream-white ribbed cardigan dress with a bow tie at the neckline, front buttons, a flowing skirt, a delicate pendant necklace, and a thin gold chain shoulder bag strap. The woman on the right has {argument name="right woman hair color" default="warm chestnut brown"} long softly curled hair, a pearl-and-flower hair clip on the side, a pale blue ribbed off-shoulder or wide-neck knit dress, a delicate necklace, a braided waist detail, and a silver chain shoulder bag strap. The background is {argument name="setting" default="a glamorous cream-and-gold boutique interior"} with glossy marble floors, built-in illuminated display shelves holding vases and decor, framed floral artwork, mirrored wall panels, warm ceiling lights, and a refined upscale atmosphere. Use natural indoor warm lighting, realistic skin and fabric texture, slight phone-camera distortion, shallow but readable background detail, soft pastel colors, and the authentic look of a casual social media snapshot. No text, no logos, no illustration style.
Infographic / Edu Visual - Japanese AI Spatial Consistency Notebook — output

Infographic / Edu Visual - Japanese AI Spatial Consistency Notebook

あいちゃんGPT Image 2
{"type":"realistic overhead photo of an open handwritten comparison notebook on a wooden desk","scene":"A squared-up top-down view of an open lined notebook spread across a warm wood table, photographed like a study note or technical evaluation memo. A right hand at the bottom right holds a blue ballpoint pen and is underlining or writing near the bottom center. Use natural daylight, soft shadows, crisp paper texture, realistic hand anatomy, and legible handwritten Japanese notes with small colored highlights.","main title":"{argument name=\"main title\" default=\"比較: GPT-Image-2 ⇔ Seedance 2.0(空間一貫性)\"}","subtitle":"{argument name=\"subtitle\" default=\"トピック: AI画像生成における空間一貫性の詳細比較\"}","layout":{"format":"two-page notebook comparison chart","sections":[{"title":"{argument name=\"left section title\" default=\"■ GPT-Image-2 の特徴\"}","position":"left page upper half","count":4,"items":["空間の整合性は○(標準的)","物体の配置(例: テーブルの上にリンゴ)","詳細な指示(例: 部屋の左側に窓)への対応力","A blue flow diagram labeled 【入力】プロンプト → 【生成】モデル処理 → 【出力】画像"]},{"title":"left results","position":"left page lower half","count":3,"items":["物体のレイアウト維持率 65%","前景・後景の一貫性 72%","複雑な指示への追従性: やや課題あり(±)"]},{"title":"{argument name=\"right section title\" default=\"■ Seedance 2.0 の特徴\"}","position":"right page upper half","count":4,"items":["空間一貫性は◎(大幅向上)","厳密な位置指定(例: 中央の棚に本)に強い","詳細な構図の維持力","A green flow diagram labeled 【入力】詳細プロンプト(空間情報含む) → 【生成】高度な空間理解モデル → 【出力】正確な画像"]},{"title":"right results","position":"right page lower half","count":3,"items":["レイアウト維持率 88%","前景・後景の一貫性 91%","複雑な指示への追従性: 非常に高い(★)"]},{"title":"center comparison","position":"bottom center across both pages","count":3,"items":["Two small framed sketch examples side by side","Left mini sketch captioned GPT-Image-2 with a small yellow warning triangle","Right mini sketch captioned Seedance 2.0 with a green check mark, separated by VS"]},{"title":"character doodle","position":"top right corner of right page","count":1,"items":["small monochrome anime girl head doodle with short bob hair, expressionless blurred face, label 『あいちゃん』 and tiny motion marks"]}],"bottom note":"空間の確さと安定性"},"visual style":"clean Japanese handwritten study-notebook aesthetic, black ink text with blue and red section headers, yellow highlighter strokes under important phrases, small blue and green boxes and arrows, neat ruled paper lines, realistic open notebook binding, slightly rounded page corners, no digital UI, no extra objects except the pen and hand.","camera":"top-down 3/4 overhead composition, notebook fills most of the frame, wood grain visible around edges, high resolution, sharp legible text","custom accent":"{argument name=\"pen color\" default=\"blue\"}"}
Comic / Storyboard - Anime Phoenix Girl Ruins Scene — output

Comic / Storyboard - Anime Phoenix Girl Ruins Scene

Naoyuki Okada / Scarlet Echoes🐲GPT Image 2
Create a cinematic anime-style fantasy illustration in a 16:9 landscape frame. A young female character, {argument name="character name" default="unnamed white-haired phoenix girl"}, stands in the foreground turned slightly left in a heroic half-body pose, with shoulder-length wavy {argument name="hair color" default="white hair"} tipped with deep red ends, a red hibiscus flower hair ornament, and a flowing red scarf. She wears a dark black jacket or cape over a white ruffled blouse and high-waisted black skirt or trousers, with warm sunset rim lighting on her clothing. Her face is intentionally obscured by a large centered square skin-toned censor block, and there is also a small semi-transparent vertical rectangular blur block near the upper left edge of the sky. Behind her, one enormous fiery phoenix wing made of red, orange, and yellow flame sweeps dramatically from her back toward the upper right, with glowing embers floating around it. The background shows a ruined stone arch and broken rocky landscape under a blue sky with soft clouds at golden hour, combining epic fantasy atmosphere, dynamic fire effects, polished digital painting, high-detail anime rendering, strong orange-blue color contrast, dramatic lighting, and a stable video-generation-friendly composition.
Comic / Storyboard - Anime Mage and Fire Dragon Ruins — output

Comic / Storyboard - Anime Mage and Fire Dragon Ruins

Naoyuki Okada / Scarlet Echoes🐲GPT Image 2
Create a cinematic widescreen anime fantasy illustration of {argument name="character name" default="a young white-haired mage girl"} standing in the foreground among ancient stone ruins at sunset. She has short, wind-tossed {argument name="hair color" default="pale white with red inner streaks"} hair, a red hibiscus-like flower hair ornament, and wears a white collared dress with a pleated skirt, a large red ribbon bow at the neck, and a dramatic black hooded cloak with deep crimson lining billowing in the wind. Her face is deliberately hidden by one flat opaque square censor block in muted salmon pink centered over the face. Behind her, a huge fiery dragon made of swirling orange-red flames arcs across the sky, its head and wing-like flame trails glowing intensely as embers scatter. The setting is a ruined fantasy landscape with broken stone arches and crumbling masonry on both sides, sparse dead branches and grasses, warm golden sunset clouds, deep blue upper sky, and strong orange rim lighting. Use a low-angle heroic composition, dynamic cloak movement, high-detail painterly anime rendering, dramatic lighting, sharp foreground character detail, atmospheric depth, and an epic magical adventure mood in a 16:9 aspect ratio.
Browse by model, style or topic

Find your next direction

ASTORIE CANVAS

Take the idea further.

Bring your prompts, references and results together in one creative space.

Open your canvas ↗