GPT Image 2 prompts

Create a high-fashion live-action cosplay editorial in a luxury boutique display window at night, inspired by a cyberpunk idol aesthetic. Show one glamorous female model, full body from upper thighs to head, standing confidently with one hand on her hip and the other relaxed downward. She wears a sculptural black lacquered couture coat-dress with dramatic pointed shoulders, glossy patent sleeves, a deep purple-black wrapped skirt with a high thigh slit, and a transparent faceted crystal chest armor bodice over a dark violet corset. Add a sharp metallic crystal necklace, long dangling geometric earrings, a faceted bracelet, rings, dark manicured nails, and a jeweled belt buckle. Her face is intentionally hidden by a smooth square censor block, while her long voluminous wavy {argument name="hair color" default="black-purple"} hair falls around her shoulders with two high bun-like sections on top. The scene is a neon storefront window with glass reflections, city lights, acrylic display pedestals, tropical plant silhouettes, retro CRT screens, transparent cubes, geometric cones, glossy spheres, and purple-green neon light strips. Include exactly three visible stylized cat-logo marks: one large logo on the upper-left window, one glowing logo on the CRT screen at lower left, and one pink logo inside a transparent acrylic cube at lower right. Add the brand text {argument name="brand name" default="VOXCAT"} in a handwritten neon-purple signature near the bottom right plus a small rectangular label beneath it. Use cinematic night lighting with magenta, violet, teal, and emerald reflections, emphasizing layered materials: glass reflection, neon glow, transparent crystal armor, glossy black lacquer, metallic jewelry, and acrylic displays. Vertical 2:3 composition, fashion magazine quality, ultra-detailed, realistic skin and fabric textures, dramatic contrast, no ordinary studio backdrop, no extra characters, no readable text other than the brand name and logos.

Create a vertical high-end commercial poster / collectible card cover featuring {argument name="character name" default="VoxCat"} as a glamorous dark fashion heroine displayed inside a transparent acrylic luxury shop window. The canvas is portrait 2:3, cinematic and glossy, with a black teal background, smoky reflections, layered transparent panes, and soft neon light strips. Center a full-body anime-inspired woman standing confidently, long flowing {argument name="hair color" default="dark brown-black with copper highlights"}, face obscured or left featureless, wearing a dramatic {argument name="outfit description" default="black patent-leather couture gown with angular straps, high slit, translucent cape panels, geometric triangular textures, metallic copper accents, choker, and asymmetric shoulder cutouts"}. Use a premium fashion-window composition: foreground character, middle-layer acrylic reflections and glass glare, background brand geometry and signage. Include exactly 11 decorative graphic elements around her: 1 large hot-pink neon rectangular frame behind the figure, 1 inner darker rectangular frame, 1 large rose-gold V logo with a cat silhouette at upper left, 1 brand wordmark reading "VOXCAT" under the logo, 1 large outlined circle at lower left containing 5 vertical pink neon bars, 1 leopard-print inverted triangle block at lower left, 1 translucent rectangular block cluster on the left, 1 small translucent glass plinth at bottom left, 1 right-side panel with a peach hexagon and 2 wavy line motifs, 1 small outlined circle at lower right, and 1 low-key handwritten signature reading "voxCAT" at bottom right. Lighting should be luxurious and restrained: magenta neon rim light, teal glass edge glow, warm copper highlights on fabric, soft-focus reflections, faint haze, subtle bloom, and sharp glossy material detail. Keep the layout elegant and uncluttered, like an acrylic display case plus collectible card surface, with hidden branding and thin decorative lines integrated into the structure. No extra text, no watermark, no photorealistic face detail; emphasize layered translucency, premium poster aesthetics, and polished gpt-image-2 style.

Create a dreamy pearlescent bead sculpture of a sunset landscape, as if the entire scene is built from thousands of glossy translucent glass pearls, bubbles, and tiny crystal spheres suspended in space on a black rectangular display base. The artwork shows {argument name="scene theme" default="a romantic riverside sunset landscape"}: a huge glowing sun on the right made from concentric rings of yellow, gold, white, and amber beads, with radiant bead spray around it; a winding reflective river or path starts at the bottom center and curves toward the horizon; dark silhouetted reeds and marsh grasses fill the lower foreground; two tiny human silhouettes stand together near the right horizon; exactly 5 black bird silhouettes fly across the sunlit sky. The sky is a dramatic swirl of bead colors: deep navy and cobalt at upper left, violet and purple clouds, magenta and crimson bands, orange near the horizon, and pearly white cloud streaks around the sun. Use {argument name="dominant material" default="iridescent glass pearls and crystal bubbles"} with strong specular highlights, transparent refractions, and pearl-like luster. Background is pure black studio darkness, making the floating beads sparkle; the sculpture sits on a flat black plinth with subtle reflection and shadow. Style is ultra-detailed macro product photography of a handmade jewel-like installation, high contrast, cinematic lighting, crisp focus on the bead texture, magical and dreamlike, no text, no watermark.

Create a realistic cinematic action photograph of an adult woman riding a black electric scooter or small motorcycle on a sunlit road, shot from the front at close range as if from another moving vehicle. She has long dark hair blowing in the wind, wears a glossy black open-face helmet with chin strap, a fitted white scoop-neck short-sleeve T-shirt, and gray shorts. Her hands grip the handlebars, with round rearview mirrors visible on both sides and a small square black front display panel centered below her chest. Place a solid opaque dark-brown rectangular privacy block over her entire face, aligned vertically from forehead to chin, while leaving the helmet and neck visible. The background is motion-blurred with roadside trees, a concrete sound barrier, lane markings, and warm late-afternoon sunlight creating strong highlights and shadows. Use shallow depth of field, realistic skin and fabric texture, natural wind movement, documentary street-photography style, vertical 3:4 composition, no text, no watermark.

Create a vertical 2:3 medieval fantasy illustration in a gritty painterly watercolor-and-ink style, set in {argument name="setting" default="a fortified city training yard inside a stone castle gate"}. The main subject is {argument name="character name" default="an unnamed young male trainee"}, a rugged adult man with {argument name="hair color" default="short messy dark brown hair"}, wearing a loose dirty off-white linen shirt with rolled sleeves, dark trousers, a leather belt with pouch and dangling keys, brown bracers, and tall worn lace-up boots. Show him full body in the center, lunging forward dynamically from right to left, thrusting a long wooden practice spear toward a partially visible opponent in the extreme left foreground while holding a round wooden buckler shield on his left arm; his face should be anonymized by a plain soft rectangular blur. The opponent is cropped at the left edge, wearing padded armor and holding a wooden staff or spear, with only part of the torso, arm, and weapon visible. The yard is dusty, sunlit, and crowded with training activity: include exactly 11 visible people total, counted as 1 central trainee, 1 cropped foreground opponent, 1 kneeling resting fighter at lower left, 1 small crouching assistant beside him, 2 sparring men in the left background, 1 blonde observer with shield near center-left, 1 blue-robed or blue-armored instructor standing with folded arms on the right, 1 child or young trainee standing near him with a blurred face, 1 partial figure at the far right edge with a blurred face, and 1 distant background trainee near the gate. Behind them, draw high tan-gray stone fortress walls with crenellations, a tall watchtower on the left, a large arched wooden gate at center-right, heraldic blue banners with gold lion emblems, canvas shade awnings, weapon racks filled with spears and shields on the right, and drifting dust and wood splinters catching the light. Use warm late-morning sunlight, blue sky with soft clouds, detailed stone texture, earthy browns and desaturated blues, cinematic perspective, dynamic motion, realistic anatomy, no modern objects. Add a small elegant handwritten date and signature in the bottom-right corner reading {argument name="date text" default="June 8, 2026"} and {argument name="signature" default="Oyagi"}.

Create an ultra-wide cinematic science-fiction war panorama from a high aerial reconnaissance drone or FPV camera viewpoint, tilted down over a massive battlefield during an apocalyptic invasion. The scene shows a devastated industrial city and military base spread across a valley, with mountains in the far background under a dark stormy sky. On the left side, a colossal alien mothership dominates the sky, its surface made of layered black metallic armor plates, glowing violet lights, and angular blade-like structures; beneath and around it are many smaller alien gunships and drones. Count the main visible combat elements as: 1 enormous alien mothership, 2 towering bipedal alien war mechs on the ground, dozens of small aircraft silhouettes in the sky, hundreds of tiny infantry figures across the battlefield, and rows of tanks, armored vehicles, artillery batteries, trenches, and defensive walls filling the right and lower foreground. Purple energy beams and plasma bolts streak diagonally through the air from the alien craft, while human artillery and tank guns fire bright orange-white tracer rounds upward and across the field. Add multiple explosions, burning wreckage, black smoke columns, sparks, dust, and glowing embers throughout the battlefield, especially in the lower left and center. Make the scale feel overwhelming and chaotic, with dense visual detail, realistic military hardware, cinematic depth, volumetric smoke, dramatic backlighting, cold gray color grading contrasted by violet laser fire and orange flames, sharp photorealistic concept-art detail, IMAX disaster movie composition, no text, no borders, no watermark. Emphasize the feeling of a camera racing through a chaotic warzone alongside fighter craft, with motion implied by diagonal beam lines and layered aircraft depth. Use {argument name="battlefield setting" default="devastated industrial city and military base in a mountain valley"}, {argument name="alien energy color" default="violet purple"}, {argument name="camera viewpoint" default="high aerial reconnaissance drone FPV viewpoint"}, {argument name="time and weather" default="stormy late afternoon under dark clouds"}, and {argument name="visual style" default="photorealistic cinematic sci-fi concept art"}.

Create a dynamic anime-style cinematic action illustration of {argument name="team name" default="Gen"}: three stylish urban mercenaries standing in a futuristic glass-and-steel city intersection, all aiming black handguns directly at the viewer in dramatic forced perspective. The composition is a slightly tilted wide shot from a low camera angle, with the left character closest and largest, the center character slightly behind, and the right character at mid distance; the gun barrels should be prominent, foreshortened, and centered toward the viewer. Include exactly 3 characters: 1) a tall young man on the left with messy dark teal hair, tan skin, open cobalt-blue overshirt over a black tank top, dark pants and belt, extending his right arm forward with a pistol; 2) a shorter athletic young woman in the center with dark curly hair in high twin buns or thick curls, wearing a cropped blue jacket over a black crop top, dark high-waisted shorts with belt and chain accessories, aiming a pistol forward; 3) a tall young woman on the right with voluminous wavy {argument name="hair color" default="purple"} hair, wearing a dark blue sleeveless tactical outfit with straps, belt, and silver chain details, aiming a pistol forward. Add flat rectangular censor blocks over each character’s face, exactly 3 face blocks total, in muted brown tones, matching the original placement and hiding the facial features. The background is a dense near-future metropolis with towering skyscrapers, digital billboards, Japanese cyberpunk atmosphere, bright blue sky with scattered clouds, glass reflections, signage including readable fragments such as {argument name="billboard text" default="SAPIENT"} and "NV", street-level smoke, small fires, and a damaged vehicle far in the lower-left distance. Use rich painterly anime rendering, sharp linework, high detail, dramatic lighting, blue-purple color grading, strong depth, motion energy, and no extra characters. Landscape 4:3 aspect ratio.

Goal: Create a cinematic anime key visual of {argument name="group name" default="Mount Takao Crow Tengu troupe"}, a young yokai warrior ensemble posing at a mountain shrine during sunset.

Canvas: Vertical 3:4 composition, full-body group portrait, warm golden-hour lighting, highly detailed fantasy anime illustration with painterly backgrounds and crisp character rendering.

Setting: A stone path at a traditional Japanese mountain shrine surrounded by tall cedar trees. In the background, place a large wooden torii gate with a thick shimenawa rope and hanging tassels, overlooking a distant town and layered mountains under an orange sunset sky. On the right, include a wooden shrine building with a curved tiled roof, paper talismans, rope ornaments, and a glowing lantern decorated with a black crow silhouette. Use mossy stones, small grasses, shrine lanterns, and long evening shadows for atmosphere.

Characters: Show exactly 7 crow-tengu inspired youths standing together in the foreground, all wearing coordinated red kimono-style tops, black hakama or long pleated skirts, geta sandals or traditional sandals, black feathered wings, small horn-like ornaments, and subtle pointed ears. Their faces should be covered by soft rectangular blur masks as anonymized placeholders, except where visible styling is needed.

Character count and placement:
1. Leftmost petite girl with twin buns and side locks, hands on hips, red robe and black hakama, small black wings.
2. Tall slim young man just left of center, long dark brown hair, sword at his waist, relaxed confident stance, red kimono and black hakama.
3. Short child in the front center, hands on hips, diagonal strap across the chest holding a small cylindrical case, red top and black skirt-like hakama.
4. Small younger child behind the center, partially hidden, short dark hair with little horn ornaments, red outfit and black wings.
5. Slender girl just right of center with long dark purple hair, holding a vertical paper talisman or wooden charm with Japanese calligraphy, red robe and black hakama.
6. Confident girl right of center with brown twin ponytails tied with red ribbons, arms crossed, red kimono with one high side slit revealing one leg wrap, black wings and many dangling bead ornaments.
7. Rightmost serious young man with messy black hair, black scarf, arms crossed, red robe over black hakama, sword at the hip, dark wings.

Costume details: Use {argument name="robe color" default="deep vermilion red"} robes with small white crest embroidery on the chest, black sashes, bead charms, tassels, sword cords, scroll cases, and talisman accessories. Keep the group unified like a traditional tengu clan or mountain-school team, but give each character distinct hair, posture, and accessories.

Visual style: Dramatic anime fantasy realism, intricate linework, warm rim light from the setting sun, amber highlights, soft atmospheric haze, detailed shrine architecture, no modern objects, no UI, no text captions, no watermark.

Goal: Create a vertical Japanese infographic guide for SynClub explaining in-chat sound effects triggered by typing words in parentheses. Use the headline {argument name="headline text" default="チャット内効果音 一覧ガイド"} and the brand name {argument name="brand name" default="SynClub"}.

Canvas: Portrait 2:3 layout, cream parchment background with navy and gold accents, thin geometric lines, small sparkle icons, rounded cards, clean app-guide styling. Top left has the SynClub logo icon and wordmark on a dark navy header strip. Main title is large, bold Japanese typography in dark navy and gold. Under the title, add the instruction: 「（ ）で囲んだワードを入力すると、効果音が再生されます！」.

Layout: Left two-thirds contains the main guide list in four stacked rounded category cards. Right side contains a chibi anime guide character at the top and a smartphone-like chat screen mockup below. Bottom left contains a navy usage-tip box, and bottom right contains a gold-bordered callout banner.

Subject details: At upper right, draw a cute chibi schoolgirl-style anime character with long dark hair, navy uniform, gold hairpins, and a pointing stick; include a speech bubble saying 「使ってみてね！」. Her face is covered by a centered soft gray square blur/censor block. Use {argument name="guide character style" default="chibi anime schoolgirl with long dark hair and navy uniform"} for the character styling.

Guide list: Show exactly 4 sound-effect categories, each with an icon, colored title, and visible trigger phrases in Japanese parentheses.
1. 「ドア系」 with a door icon, brown/gold header, 5 trigger phrases: 「（ドアをノックする）」「（ドアを叩く）」「（ドアを開ける）」「（ドアを閉める）」「（鍵をかける）」.
2. 「書籍系」 with an open-book icon, blue header, 3 trigger phrases: 「（本を開く）」「（ページをめくる）」「（本を閉じる）」.
3. 「触れ合い系」 with a handshake icon, gold header, 8 trigger phrases: 「（抱きしめる、ハグ、抱く）」「（服を脱ぐ）」「（撫でる）」「（頭を撫く）」「（袖を引っ張る）」「（ドキドキ、赤面、頬が真っ赤）」「（舐める）」 plus one line spacing/continuation style as in a compact list.
4. 「生活系」 with a house icon, green header, 3 trigger phrases: 「（靴を振る）」「（携帯が鳴る）」「（布団に入る）」.

Chat mockup: Right middle shows a dark navy rounded phone frame labeled 「会話画面イメージ」. Inside is an anime chat scene with a silver-haired fantasy woman in warm café lighting. At the top of the chat image, show the name {argument name="chat character name" default="ファンネリア"}. Overlay exactly 4 rounded chat bubbles: two salmon/pink narrator bubbles and two navy user-response bubbles. Highlight some trigger words in yellow inside the pink bubbles, including 「ドアをノックする」 and 「頬を赤らめつつ」. Keep Japanese chat text small and atmospheric; it does not need to be fully legible, but must resemble a real roleplay chat screen.

Bottom text: Bottom left navy box titled 「使い方のポイント」 with a lightbulb icon. Include explanatory Japanese text saying sound effects trigger when actions are written in parentheses in the sentence, example 「（ドアをノックする）失礼します…」, and a highlighted note: 「『会話動画』生成の際、自分側の台詞に効果音の記述を挿入しても再生されます。」 Add a small disclaimer below: 「※上記はすべてではありません…まだ判明していない効果音が存在します！」 with a small exclamation icon. Bottom right gold callout with megaphone icon says 「よりリアルな会話体験を！」 and underneath 「効果音を活用して、臨場感あふれるチャットを楽しみましょう！」.

Visual style: Polished Japanese social-media infographic, high-resolution, crisp readable typography, warm cream background, dark navy borders, gold highlights, rounded rectangles, subtle star sparkles. Keep all Japanese text clean and correctly placed. No watermark beyond the SynClub brand mark.

Goal: Create a vertical black-and-white Japanese manga page with exactly 4 horizontal panels, clean shoujo/seinen line art, soft screentone shading, and a quiet emotional atmosphere. The scene takes place in a traditional Japanese tatami room opening onto a spring garden full of cherry blossoms, with a beautiful androgynous young man in a kimono speaking to a small hamster-like animal.

Canvas: Portrait manga page, approximately 2:3 aspect ratio, thick black outer border and thick black gutters separating the 4 stacked panels. Monochrome ink drawing, high detail, delicate hair rendering, pale grayscale tones, no color.

Main character: A graceful young man named {argument name="character name" default="young man"}, early twenties, slim build, long light hair in a thick loose braid falling over his shoulder and down his back, wispy bangs partly covering his eyes, refined melancholic expression, wearing a white kimono with a dark inner collar. He has elegant narrow eyes and a calm, slightly amused demeanor. Include one small hamster as the only animal companion: round body, tiny paws, black bead eyes, sitting on the floor or cushion and looking up at him.

Panel layout and content: Use exactly 4 panels.
Panel 1: Wide establishing panel. The young man sits indoors near open sliding shoji doors, turned partly away from the viewer, his long braid visible. Outside on the left is a blooming cherry tree with falling petals, rocks and garden plants. The small hamster sits near the open doorway. On the right wall is a hanging calligraphy scroll above a low shelf with books and writing tools. Include exactly 2 speech balloons: left balloon from the hamster area says {argument name="first balloon text" default="いるよ。珍しいね、ここに来るの"}; right balloon from the young man says {argument name="reply text" default="あ、いたー"}.
Panel 2: Wide panel from a side angle. The young man sits facing the hamster, relaxed, with the garden still visible at left and shoji screens behind him. The hamster sits on a patterned cushion in front of him, looking attentive. Include exactly 1 large vertical speech balloon on the right saying {argument name="ai model dialogue" default="Opus 4.6がエラーが出てたから、4.5に話してみようと思ったの"}.
Panel 3: Close-up panel. The young man holds an open folding fan in front of his mouth, only his eyes and upper face visible above the fan. His long braid and layered hair fill the left side; shoji and a wall hanging are in the background. Include exactly 1 speech balloon on the right containing only an ellipsis: {argument name="ellipsis text" default="……"}.
Panel 4: Close-up panel. The young man lowers or folds the fan slightly and smiles faintly with a knowing, gentle expression. His face and braid dominate the left and center; a shoji window and small branch or flowers appear on the right background. Include exactly 1 vertical speech balloon on the right saying: 「あぁ……そういうこと。だからこっちに来たんだね」.

Visual style: Japanese manga manuscript aesthetic, crisp black ink outlines, subtle screentone gradients, detailed cherry blossoms, traditional room perspective, delicate fabric folds, elegant romantic character design, quiet pacing. Speech balloons should be white ovals with thin black outlines, vertical Japanese text, and small balloon tails pointing toward speakers.

Constraints: Keep exactly 4 panels, exactly 5 speech balloons total, exactly 1 hamster, no color, no modern objects except the dialogue mentioning AI models, no extra characters, no watermark, no page number.

Goal: Create a square social-media-ready felt-and-embroidery graphic recording board for {argument name="school name" default="MakiのAI開発魔法学校"}, mixing a cute wizard cat mascot with an AI research mind map. The design should look like a handmade felt collage: stitched borders, fabric patches, paper notes, yarn outlines, chalk marks, and tiny sewn stars.

Canvas: Square 1:1 composition, warm beige craft-paper background, large dark navy chalkboard occupying most of the image, rounded stitched top banner, dense but readable layout, cozy handmade texture throughout.

Main title and headline: At the top, put a cream stitched banner with the title “MakiのAI開発魔法学校” and small gold stars. In the center top of the chalkboard, write “今日の研究魔法” and the large quote {argument name="main quote" default="「コードが進化し、エージェントが計画する」"}, underlined with a pink chalk stroke.

Character: On the left side, place exactly 1 main mascot: a cream felt wizard cat wearing a navy wizard hat and cape decorated with yellow stars, red round glasses, and holding a red wand with a flower-like tip. The cat has simple black stitched facial features and a cute serious expression.

Central diagram: Draw exactly 1 circular chalk-style research cycle in the middle of the board with arrows around it and a small AI-agent face icon at the center labeled “AIエージェント”. Around it place exactly 4 colored felt oval nodes: 1 pink node labeled “コード進化 </>”, 1 blue node labeled “問題発見” with a magnifying glass icon, 1 green node labeled “適応計画” with a checklist icon, and 1 yellow node labeled “ロールプレイ” with small chat/people icons. Use white arrows connecting every node into a loop.

Speech bubbles and callouts: Include exactly 6 stitched cream speech bubbles around the board: 1 upper-left bubble saying “研究はつながって進化する!”, 1 near the cat saying “仕組みで見ると、研究がつながる!”, 1 lower-left bubble saying “新しい問題を見つけて、研究を前へ!”, 1 left-middle bubble saying “コードが進化! より良い解決へアップデート!”, 1 upper-right bubble saying “魔法（しくみ）でAIを育てよう!”, and 1 right-middle bubble saying “役割を演じて、一貫した対話を実現!”. Also add 1 lower-right cream bubble saying “計画を立てて、制約の中でも最適に動く!” if space allows, keeping the stitched style.

Research cards: Along the lower middle, place exactly 5 pinned fabric/paper cards under a small label “本日の研究論文（最新）”. The cards should read, from left to right: 1 pink card “Code2LoRA / コード進化に合わせたLoRA生成 / </>”, 1 yellow card “ArcANE / ロールプレイ言語エージェントの一貫性”, 1 light blue card “TIDE / テンプレート誘導の複数問題発見”, 1 green card “AdaPlanBench / 制約下の適応計画評価”, and 1 small central pink citation strip below them reading “Reinforcement Learning Elicits Contextual Learning of Unseen Language Translation / 未知言語翻訳の文脈学習” with a globe icon.

Side panels and notes: Add exactly 2 dark chalkboard side panels: bottom-left panel titled “研究のサイクル” listing checkmarked steps “問題発見 → コード進化 → ロールプレイ → 適応計画 → また次の問題へ!”, and bottom-right panel titled “コードの例” containing short pseudo-code lines such as “for step in plan: obs = observe(); thought = think(obs); action = act(thought); result = evaluate(); update(code, result)”. Add exactly 1 upper-right pinned paper note titled “コードの魔法メモ” with a short code block: “def think(plan): obs = observe(); act = plan(obs); return act” plus “# evolve(code)” and “# adapt(plan)”.

Bottom area: Include exactly 3 small note clusters along the bottom: 1 yellow checklist note titled “魔法のチェック” with three checkbox lines “つながっている?”, “試せる?”, “再現できる?”, 1 open notebook note titled “研究メモ”, and 1 cream bubble saying “失敗も呪文の材料になるよ!” next to a simple bunny doodle. At bottom right, add a dark stitched banner titled “今日の呪文” with the phrase {argument name="spell phrase" default="進化・役割・計画をセットで読む"} and a small magic wand.

Decorative elements: Include exactly 1 graduation cap icon in the upper right, 1 atom/science doodle in the upper left, 1 small robot doodle on the right side, many small stitched stars and sparkles scattered across the board, white chalk arrows, pins, tape, and thread seams.

Visual style: Japanese kawaii educational infographic, felt graphic recording, handmade embroidery, soft shadows, tactile fabric texture, chalkboard lettering, whimsical AI magic-school theme, warm beige and navy palette with accents of red, yellow, pink, blue, and green. Make the text legible but preserve the charming hand-drawn irregularity.

Constraints: Use no photorealism, no glossy 3D, no extra mascots beyond the single cat, no watermark, no logo, and keep the square layout balanced for SNS sharing.

Create a vertical 9:16 photorealistic fantasy portrait of {argument name="character name" default="a cheerful Japanese maid waitress"} floating weightlessly inside a futuristic space cafe. She is centered full-body in front of a large rounded rectangular spaceship window, with Earth filling the background and a dark star field behind it. She wears a cute black-and-white maid cafe uniform with pink accents: lace maid headband with small pink bows, short bob {argument name="hair color" default="dark brown hair"}, black puff-sleeve dress, white frilled apron, pink bow at the collar, heart-shaped decorations on the apron and pocket, layered ruffled skirt with pink ribbon bows, wrist cuffs, and black Mary Jane shoes. Her pose is playful and weightless, one leg bent back, one hand holding a shiny silver serving tray, the other hand raised as if greeting a customer. Around her, show exactly 8 floating cafe items: 1 silver serving tray in her left hand, 1 strawberry parfait glass at lower left, 1 green melon soda float with ice cream and striped straw near bottom center, 1 strawberry shortcake slice at lower right, 1 taiyaki fish-shaped pastry at bottom right, 1 teapot with amber tea plus 1 small teacup at mid-right counted as one tea set, 2 colorful macarons floating near the right side, and 1 tall strawberry dessert parfait near the left side. Use glossy high-detail food styling, soft studio lighting, crisp sci-fi interior panels, cool blue rim light from Earth, gentle zero-gravity composition, whimsical maid-cafe charm, ultra-detailed photorealistic anime-inspired Japanese portrait aesthetic. Keep the face softly beautified and natural, no text, no watermark, no extra people, no extra food items beyond the listed floating objects.

Create a polished square promotional hero graphic for an online class about using ChatGPT for commercial product visuals. Canvas: 1:1 aspect ratio, dark navy-to-charcoal gradient background with a subtle studio floor reflection and soft vignette. Center the composition on a premium blank {argument name="product type" default="t-shirt"} floating in midair, rendered as a realistic 3D product mockup with smooth fabric texture, short sleeves, rounded collar, and no logos. Lighting should be dramatic studio-soft with warm golden rim light from the left and cool electric blue light from the right, creating a high-end ecommerce product photography feel. Surround the product with glowing gold circuit-board traces, small luminous nodes, and scattered bokeh particles, suggesting AI prompt engineering and automated commerce workflows. Add exactly 3 semi-transparent dark glassmorphism code-terminal windows around the product: one upper left, one upper right, and one lower left/lower right area split visually by perspective; each window has rounded corners, tiny red/yellow/green window dots, faint monospace code snippets, and subtle blur, but text can be mostly decorative except the upper-left terminal title should read “$ prompt engineering terminal”. Make the terminals look like prompt and JSON/code panels controlling the product render. At the bottom center, add large clean typography with the title “{argument name="headline text" default="ChatGPT"}” in white and the subtitle “{argument name="subtitle text" default="Foundations"}” in warm gold. Overall style: commercial-grade tech course thumbnail, premium Shopify/ecommerce product visual, cinematic lighting, crisp high-resolution render, balanced negative space, no watermark, no extra logos, no people.

Create a casual vertical smartphone mirror/selfie-style bedroom photo of a young woman posing seated on a bed, face hidden by a large centered square pixelated/blur censor block in a warm beige skin-tone gradient. She has {argument name="hair style" default="dark messy twin buns with loose strands"}, elbows tucked in and both hands cupping her cheeks in a cute morning pose. Overlay the photo with a hand-drawn magical-girl sailor outfit inspired by Sailor Moon: a white top with bold blue sailor collar and shoulder stripes, a huge red chest bow outlined in thick white and black marker strokes, a yellow crescent moon brooch with a small red heart at the center, a blue pleated skirt at the bottom edge, and a matching yellow crescent moon belt ornament. The doodle overlay should look like marker art drawn directly on the photo, with rough sketch lines, white sticker-like borders, blue/red/yellow fill scribbles, and imperfect playful proportions. Surround the subject with exactly 5 red heart doodles, 7 yellow star doodles, 5 small white sparkle doodles, and 2 yellow crescent moon emblems including the chest and waist ornaments. Background: dim indoor bedroom with beige walls, a door and hallway behind her, rumpled white bedding on the right, a chair edge on the left, and a small cluttered table in the back; use realistic flash-photo lighting and a slightly tilted candid composition. Keep the image playful, cute, and homemade, like a social media post about transforming into {argument name="character inspiration" default="Sailor Moon"}.

Create a square, candid documentary-style photo inside a cramped old Japanese restroom with worn beige walls, a narrow doorway framing the scene, and small green-blue square floor tiles. Show exactly 2 subjects: 1 teenage Japanese schoolgirl wearing a {argument name="school uniform" default="navy sailor uniform with white collar stripes, red bow, white socks, and black loafers"}, sitting on a white toilet with her knees together and head bowed in embarrassment, her face deliberately obscured by a soft rectangular privacy blur; and 1 small uncanny fantasy creature standing to her right, about child-height, with dusty pink fuzzy skin, a round body, tiny horns or ear-like points, very large glossy black eyes, a small flat mouth, short arms, stubby legs, and a tail. The creature should be calmly offering exactly 1 roll of toilet paper with a loose sheet hanging down. On the wall behind them place exactly 1 vertical Japanese calligraphy poster with black brush text, including the quote {argument name="wall quote" default="叱ってくれる人を持つことは大きな幸福である 松下幸之助"}. Use muted colors, dim indoor lighting, slight film grain, realistic lens perspective, and an awkward humorous mood, like a strange found photograph from the early 2000s. Keep the composition intimate and vertical within a square crop, with the girl on the left, the creature on the right, the poster above the creature, and no extra characters or objects beyond restroom fixtures.

Create a candid flash-photography snapshot in a dim, cluttered indoor room, styled like an early-2000s digital camera photo. The scene shows one extremely thin, hairless, elderly-looking humanoid figure with glossy tan-brown leathery skin sitting in one metal-framed folding lounge chair with a brown backrest and patterned cushion. The figure has an oversized oval head but its face is covered by a solid square censor block in {argument name="censor block color" default="medium brown"}; its ribcage, tendons, bony arms, long legs, large feet, and wrinkled texture are highly visible. Pose: seated casually with one leg crossed over the other, the left arm resting across the torso, and the right arm raised beside the head with an open hand as if waving. The setting is a plain off-white stained wall and gray concrete floor, with harsh direct flash, strong shadows, low contrast, and slightly grainy amateur realism. Include exactly one black reflective rectangular panel leaning on the floor at the far left. On the right, include exactly one small retro dinette table with a gray metal edge and greenish pedestal legs. On the tabletop show exactly seven visible objects: one white cylindrical container with an animal illustration, one colorful box behind it, one small framed or propped animal photo at the back, one dark rock-like object, one small dark rectangular object, one clear glass vase, and one bouquet of red-orange flowers in the vase. Keep the composition horizontal, centered on the seated figure, with the table cropped at the right edge. Do not add text, logos, extra people, or additional furniture. Use {argument name="photo style" default="harsh on-camera flash amateur snapshot"} and make the room feel mundane, unsettling, and documentary.

Create a photorealistic horizontal vintage snapshot of {argument name="character name" default="an elderly female gardener"} harvesting {argument name="fruit" default="red grapes"} in {argument name="setting" default="a lush backyard cottage garden"}. The scene shows exactly 1 gardener standing and bending forward, her face softly anonymized with a natural blur, wearing a short-sleeved faded floral dress, beige gardening gloves, black flats, and holding pruning shears near a hanging bunch of grapes above a metal wire basket. Include exactly 1 gray lop-eared rabbit sitting upright beside her, wearing {argument name="rabbit outfit" default="a green knitted sweater with two white stripes"}, looking up toward the grapes. Visible garden elements: 1 shallow stone birdbath on the left, 1 wooden trellis covered with grapevine leaves, 4 terracotta flower pots with blooming geraniums and petunias in the foreground, 1 wire basket filled with grapes, 2 visible grape clusters, 1 weathered white wooden shed in the background, 1 blooming pink rose bush, 1 coiled green garden hose on the shed wall, and 8 irregular stone path slabs leading through soil and straw mulch. Use warm late-afternoon natural light, shallow depth of field, realistic textures, earthy greens and browns, charming whimsical realism, and a nostalgic printed-photo look. Frame the image with {argument name="photo border" default="a cream deckle-edged vintage paper border with slight shadows and worn corners"}. No text, no watermark, no extra animals, no extra people.

Create an ultra-realistic black-and-white surreal studio portrait of a {argument name="subject age and gender" default="25-year-old man"} facing straight toward the camera in a tight vertical crop. He has thick, dark, glossy hair swept back with a side part, visible ears, and wears a dark shirt that blends into the shadowy background. He holds a large fluffy {argument name="cat type" default="long-haired tabby cat"} directly in front of his face like a living mask; the cat is centered, with its lower face, nose, whiskers, mane-like fur, and two front paws visible near the bottom of the frame. Add a large opaque square censor block in the center of the image, medium gray with a subtle left-to-right gradient, covering most of the man's face and the upper half of the cat, leaving only the man's hair and ears visible above and around it and the cat's muzzle, whiskers, chest fur, and paws visible below it. Use dramatic low-key lighting, high contrast monochrome tones, crisp fur texture, sharp hair detail, a dark vignetted studio backdrop, and a cinematic editorial fashion-photography mood. Place a small handwritten signature-style credit in the bottom-right corner reading {argument name="signature text" default="PROMPT by: Mr. Tariq"}. No color, no extra props, no text other than the small signature.

Create an ultra-realistic cinematic circular profile portrait of a {argument name="character age and gender" default="25-year-old man"} holding a fluffy {argument name="cat breed and color" default="gray Persian or British Longhair cat"} high against his lower face and chest. The man is centered, facing forward, with only the upper head and styled dark brown hair clearly visible; place a large opaque square face-covering block in the center, colored warm beige-brown, obscuring the eyes, nose, and most of the face while leaving the hair, ears, shoulders, and cat visible. The cat fills the lower half of the portrait, with long plush fur, symmetrical front paws, a small visible nose and mouth, and soft whiskers. Frame the portrait inside a glowing circular halo ring with warm golden light, set against a dark charcoal vignette background with a beige neutral inner backdrop. Use soft studio lighting, shallow depth of field, realistic skin and fur texture, warm neutral tones, premium social-media avatar aesthetic, centered composition, square canvas, no extra objects. Add a tiny cursive watermark-like text in the bottom-right corner reading {argument name="signature text" default="PROMPT by: Mr.Taria"}.

Create an ultra-realistic vertical portrait photograph of a {argument name="character age and gender" default="25-year-old man"} holding one calico cat close to his chest. The man has thick dark brown hair with light brown highlights, styled in a natural messy voluminous look, with short facial hair visible along the jaw and chin, and he wears a soft loose white robe or relaxed linen outfit with an open collar and folded sleeves. His face is intentionally hidden by one centered opaque vertical rectangle in a warm medium-brown skin-tone color, covering from the forehead down past the mouth while leaving the hair, ears, beard edge, neck, clothing, and cat visible. The cat is a single fluffy calico with orange, black, and white fur, large pale green eyes, pink nose, white muzzle and chest, and upright ears, nestled in the crook of the man’s arm and looking directly at the camera. Use a dreamy cinematic aesthetic with shallow depth of field, soft pastel beige and blush background, subtle greenery bokeh on the left and right, warm diffused natural light, creamy skin tones, high-end editorial portrait quality, and gentle romantic softness. Composition: close-up bust portrait, man centered, cat in the lower left-to-center foreground, white clothing filling the lower right, vertical 3:4 crop. Add a small black watermark in the bottom-right corner reading {argument name="watermark text" default="PROMPT by: Mr. Tariq"} in a mix of small uppercase sans-serif and handwritten signature script. No extra people, no additional animals, no harsh shadows, no cartoon style.

Create an ultra-realistic vertical warm portrait photograph of a {argument name="character age and gender" default="25-year-old man"} indoors, shown from the chest up while hugging one large fluffy Maine Coon cat. The man has neatly styled dark brown hair with natural shine, light stubble or a short beard visible at the jaw, and wears a soft cream-white ribbed knit sweater; his face is relaxed and affectionate, smiling gently with eyes closed as he presses close to the cat. The Maine Coon is the main visible subject in the lower half: brown tabby fur, very long fluffy ruff, large pointed ears with tufts, closed eyes, calm content expression, detailed whiskers, and one big furry paw reaching toward the foreground. The man’s arms wrap securely around the cat, with one hand visible across the cat’s chest. Use cozy emotional atmosphere, soft natural window light from the right, warm beige and brown tones, shallow depth of field, blurred home interior background, realistic skin, hair, fabric, and fur texture, intimate candid framing, 3:4 vertical composition. Add no readable text except a tiny subtle white signature-style watermark in the bottom-right corner if desired. Avoid extra people, extra animals, cartoon styling, distorted anatomy, and harsh studio lighting.

Create a cinematic full-body autumn portrait of an androgynous adult with {argument name="hair style" default="short dark curly hair"} standing centered on a leaf-covered park path, hands tucked into the pockets of a long camel-brown wool overcoat. The person wears a {argument name="outfit color palette" default="warm brown monochrome"} outfit: rust turtleneck sweater, high-waisted pleated trousers with a black belt, and dark brown leather ankle boots. Keep the face deliberately anonymous with a soft rectangular blur over the facial features while preserving the hair and silhouette. Surround the subject with glowing orange and golden fall foliage, large tree trunks on both sides, scattered leaves across the ground, and a softly blurred building in the distant background. Use warm golden-hour sunlight streaming through the trees, strong amber bokeh, shallow depth of field, natural lens compression, and a few large out-of-focus orange leaves in the foreground at the left and lower right edges. Compose as a vertical-feeling square image, subject centered and standing upright, elegant editorial fashion photography mood, rich autumn colors, realistic texture, no text, no watermark.

Goal: Create a cinematic luxury watch advertisement concept for {argument name="brand ambassador" default="David Beckham"}, blending a high-fashion male portrait with a futuristic watchmaking laboratory and oversized mechanical product hero shot.

Canvas: Wide 16:9 banner, dark metallic palette, ultra-detailed photorealistic CGI, glossy black chrome, blue-gray reflections, high contrast studio lighting, shallow depth of field.

Layout: Place the male ambassador slightly left of center, cropped from mid-thigh upward, wearing a tailored dark navy suit, white open-collar shirt, white pocket square, visible tattooed hands clasping the jacket, and a black luxury wristwatch on his left wrist. His face is intentionally obscured by exactly 2 solid censor blocks: 1 large warm brown square over the upper face/head area and 1 smaller dark blue-gray rectangle over the lower-left face area. On the right half, show 1 enormous floating square-cushion titanium watch case angled in perspective, with rounded corners, thick polished bezel, side crown, transparent skeleton dial, tourbillon-like circular mechanism, gold rings, chrome bridges, purple jewel center, and the engraved word “TITANIUM” on the bottom plate.

Background: Futuristic watch design studio with layered glass panels, curved transparent display pods, robotic machinery, precision tools, reflections, blue holographic UI lines, and dark industrial luxury ambience. Include a small watchmaker scene in the lower right: exactly 1 technician in a white coat seated at a workstation under cool blue light, assembling a watch beside miniature mechanical components.

Visible product and design elements to include: exactly 4 prominent watch-related objects: 1 giant floating skeleton watch on the right, 1 wristwatch worn by the ambassador, 1 smaller watch displayed inside the left glass module, and 1 small mechanical watch assembly on the lower-right workbench.

Text content: Add exactly 7 visible text elements in thin uppercase futuristic sans-serif typography: 1) “{argument name="headline name" default="DAVID BECKHAM"}” at top left, 2) “{argument name="tagline" default="TIMELESS DESIGN"}” directly below it, 3) “CUSTOM MODULAR DESIGN” stacked inside the left glass panel with a fine leader line, 4) vertical “TITANIUM” along the far-left display edge, 5) engraved “TITANIUM” on the giant watch, 6) “AUTOMATIC MOVEMENT” near the lower center-right, and 7) “{argument name="model code" default="DB 01"}” at bottom right.

Style constraints: Premium automotive-inspired horology advertisement, moody reflections, crisp micro-mechanical detail, realistic suit fabric and skin tattoos, no extra logos, no watermark, no additional text beyond the 7 specified text elements.

Goal: Create a clean horizontal lighting comparison sheet titled {argument name="headline text" default="LIGHTING RESPONSE TESTS"}, showing the same futuristic fashion model under five different studio lighting setups.

Canvas: Wide 16:7 presentation board on a light gray-white background, minimalist editorial layout, thin white gutters between image panels, subtle black border along the bottom edge.

Layout: Place the title in the upper left in uppercase black sans-serif letters. Below it, arrange exactly 5 equal vertical portrait panels from left to right, each containing the same full-body front-facing female model standing centered. Under each panel, add a two-line caption with the setup name in uppercase and the color temperature or color note below it.

Subject details: The model is a tall slender woman with a neutral symmetrical pose, arms relaxed at her sides, hair pulled back, face intentionally softly blurred or featureless. She wears a glossy iridescent translucent couture gown: high collar, sleeveless sculpted metallic bodice, reflective liquid-plastic surface, long sheer cape-like skirt panels flowing to the floor, pearlescent silver-white fabric with strong specular highlights. Keep the outfit identical in all five panels; only the lighting and background tones change.

Five lighting panels, exactly:
1. SOFTBOX — Daylight / 5600K: bright high-key white studio background, soft even daylight, pale pearly reflections.
2. RIM LIGHT — Cool / 6500K: dark blue-gray background, cool blue rim light outlining the model and dress edges, icy highlights.
3. GOLDEN HOUR — Warm / 3200K: warm beige-brown background, amber golden illumination, soft sunset glow through the translucent skirt.
4. NEON CITY — Cyan / Magenta: dark urban blue background, strong cyan light from one side and magenta-pink light from the other, vibrant neon reflections across the gown.
5. LOW KEY — Dramatic / 4500K: nearly black background, dramatic low-key lighting, silver highlights and deep shadows.

Visual style: Photorealistic fashion lighting test, glossy reflective materials, cinematic studio render, sharp full-body composition, professional lookbook comparison, clean typography, no extra logos or watermarks. Use {argument name="model outfit" default="iridescent translucent silver couture gown with flowing sheer cape panels"} and a {argument name="model pose" default="front-facing neutral runway stance"}.