GPT Image 2 prompts

Create a highly realistic candid portrait photo of a cheerful {argument name="subject" default="Japanese high school girl"} in a bright classroom, wearing a traditional sailor-style school uniform: white short-sleeve blouse, navy sailor collar with white stripes, large burgundy bow at the chest, and navy pleated skirt. She has {argument name="hair style" default="long black hair in two loose braids with wispy bangs"}, round thin metal eyeglasses, clear youthful skin, and a warm open-mouthed smile as she turns her head to the right while leaning forward with both hands on a desk. The scene includes exactly three other visible classmates in the background and foreground: one blurred girl seated by the left window touching her face, one blurred girl on the right edge in the same uniform, and one dark-haired student seen only from behind in the lower left foreground. Set the classroom in soft morning or afternoon sunlight streaming through large windows on the left, with a green chalkboard and indistinct posted paper in the back, wooden desks, notebooks, and a shallow depth of field. Use natural backlighting, gentle rim light on the hair, realistic skin texture, crisp focus on the main subject’s face, softly blurred background, 85mm lens look, documentary school-life photography, vertical 3:4 composition, no anime stylization, no text, no watermark.

Goal: Create a dramatic vertical promotional poster for {argument name="character name" default="hirokaji"}, presenting them as an expert in advanced prompt design and AI creativity.

Canvas: Portrait poster, approximately 3:4 aspect ratio, high-resolution, cinematic Japanese tech-advertising style. Use a deep navy-to-black background with electric blue highlights, glowing particles, lens flares, and network-like lines.

Layout: The poster has 4 main visual zones: 1) a huge oversized Japanese headline filling the top and upper background, cropped by the canvas edges; 2) a realistic anthropomorphic dog portrait occupying the lower-left foreground; 3) a magical AI/knowledge scene with books and light trails across the middle-right; 4) name and handle typography at the lower-right.

Text content: Use a massive bold white Japanese headline in the background, extremely condensed and poster-like, reading {argument name="headline text" default="プロンプトを知の設計図に"}. Place a smaller white Japanese tagline near the center-right reading {argument name="tagline text" default="AIの可能性を、誰でも再現できる形へ。"}. At the bottom-right, set the name in very large lowercase white Latin letters: {argument name="display name" default="hirokaji"}. Directly below it, add the handle in smaller bold white text: {argument name="social handle" default="@hirokaji_"}.

Subject details: Create exactly 1 dog character in the lower-left foreground: a realistic Jack Russell terrier with a white face, tan ears and tan patches, dark expressive eyes, black nose, looking straight at the viewer with a serious intelligent expression. Dress the dog like a professor or prompt engineer: exactly 3 visible clothing pieces — a brown tweed blazer, a crisp white dress shirt, and a navy blue bow tie.

Objects and effects: Include exactly 2 open floating books on the right side, one larger open book near the upper-right and one smaller open book near the mid-right. Include exactly 1 stack of books at the far right edge near the lower-right, made of 4 visible books. Include exactly 1 black pen floating diagonally near the lower-middle-right. Include exactly 2 loose white paper sheets flying near the right side. Add a bright blue-white burst of light near the center-right, connected to thin glowing network lines, star points, and sweeping blue energy waves.

Visual style: Photorealistic dog and clothing combined with a glossy cinematic sci-fi poster background. High contrast, sharp focus on the dog, bold white typography, intense electric blue lighting, dynamic diagonal motion, professional Japanese magazine-cover energy.

Constraints: Keep all typography white and bold. The huge Japanese headline must dominate the top half and be partially cropped for impact. Do not add extra people, extra animals, logos, watermarks, or additional readable text beyond the specified headline, tagline, name, and handle.

Create a high-quality vertical outdoor portrait photograph of {argument name="character name" default="a young South Asian man"} sitting relaxed on an ornate black wrought-iron park bench in a garden. He has medium warm skin tone, natural skin texture, thick black hair swept back with volume, dark expressive eyes, and a neatly trimmed full beard and mustache. He wears a dark chocolate-brown traditional kurta shalwar with a short stand collar and button placket, sleeves casually rolled to the forearms. His pose is calm and confident: seated slightly angled, one hand resting on his lap showing a simple silver ring and black beaded bracelet, the other hand lifted to his hair, showing a gold wristwatch. He looks directly at the camera with a serious, composed expression. Behind him is a lush flowering bougainvillea-covered arbor with vivid magenta and white blossoms, rendered with creamy shallow-depth-of-field bokeh. Use warm golden-hour sunlight from the side/back, soft natural shadows, realistic skin detail, cinematic color grading, 85mm portrait lens look, crisp subject focus, and a premium lifestyle photography feel. Vertical 9:16 composition, waist-up to full seated crop, no text, no watermark, no extra people.

Create a hyper-realistic cinematic action photo of a gigantic police humanoid robot stopping a runaway box truck in the middle of a narrow Japanese city street. The scene shows one towering bipedal mech labeled with the Metropolitan Police Department markings, using both massive mechanical hands to crush and hold the front cab of one dirty white delivery truck; the truck’s hood and windshield area are dented inward, the front wheel area is damaged, and dust or tire smoke rises from the asphalt. The robot is white, black, and dark gray with many exposed mechanical joints, armor plates, bolts, hydraulic details, caterpillar-like modules, red emergency light bars on both shoulders, and white Japanese text plus “MPD” on two black shoulder panels; its head is a horizontal rectangular sensor/cannon block with multiple circular lenses facing forward. Set the action on a bright sunny day under a clear blue sky, surrounded by tall apartment buildings, dense overhead utility wires, power poles, street signs, sidewalks, and a few distant pedestrians for scale. Use a low street-level camera angle looking upward to emphasize the robot’s huge scale, vertical 4:5 composition, sharp realistic lighting, hard midday shadows, detailed urban clutter, gritty dust, damaged metal, documentary news-photo realism, no illustration style, no cartoon effects. Include exactly two main subjects: {argument name="robot label" default="MPD"} police mech and {argument name="vehicle type" default="white box truck"}. The visible text on the robot should include {argument name="Japanese police text" default="警視庁"} and “MPD”; avoid extra readable text except incidental street details.

Goal: Create a realistic Instagram Reels profile grid showing AI fitness influencer nutrition content, with 12 separate vertical thumbnails arranged in a 3-column by 4-row grid. The overall look should be warm, muted, lifestyle-oriented, and social-media ready, like a screenshot of a creator’s reels tab.

Canvas: Portrait screenshot-style composition, 3:4 aspect ratio. Divide the canvas into exactly 12 equal vertical reel tiles with thin white gutters between them. Each tile has a small white Instagram Reels icon in the top-right corner and a white play-triangle icon with a view count in the bottom-left corner.

Subject: The same fit young woman appears throughout the grid, an AI fitness influencer with a lean athletic body, long dark hair usually tied in a loose ponytail or bun, often taking mirror selfies with a phone covering her face. She wears minimal black, brown, olive, or cream sports bras and high-waisted leggings. Use {argument name="influencer style" default="minimal dark-haired fitness influencer in neutral athleisure"}. Keep her recognizable across all thumbnails while varying pose, outfit color, and room.

Visual style: Photorealistic iPhone lifestyle photography, soft natural window light mixed with warm indoor lighting, beige and taupe interiors, clean modern apartments, subtle grain, slightly dim cozy exposure, muted contrast, aspirational wellness aesthetic. No cartoon or illustration style.

Layout and exact tile count: Include exactly 12 reel thumbnails:
1. Top-left: mirror selfie in a minimal room with potted tree and exercise bike, black sports bra and black leggings, seated pose. Text: “What I eat to feel full longer”. Views: “3,512”.
2. Top-center: woman in modern kitchen drinking from a mug, brown sports set, bowl of salad and drink on counter. Text: “People think my meals are boring...”. Views: “98.7K”.
3. Top-right: standing mirror selfie in a simple room, black sports bra and black leggings. Text: “How I actually stay lean”. Views: “256K”.
4. Second row left: mirror selfie sitting on floor near beige sofa in bright living room, dark olive sports set. Text: “pov: you stop obsessing over calories”. Views: “112K”.
5. Second row center: back view cooking at a stovetop in a warm modern kitchen, black strappy sports bra, vegetables on counter. Text: “Where is the protein?”. Views: “41.3K”.
6. Second row right: seated at a dining table with bowl of food and glass, cream sports bra, long dark hair down. Text: “What I eat to avoid overeating”. Views: “193K”.
7. Third row left: gym mirror selfie in dim home gym, olive leggings and dark sports bra, cable machine in background. Text: “This is how I eat every day”. Views: “146K”.
8. Third row center: bedroom mirror selfie, cream sports bra and brown leggings, bed and soft lamp behind her. Text: “‘omg I could never eat like that’”. Views: “27.8K”.
9. Third row right: back view cooking in kitchen, dark sports bra and leggings, pot on stove and cutting boards nearby. Text: “Why don’t you just eat normal food?”. Views: “66.2K”.
10. Bottom-left: seated mirror selfie in cozy bedroom or living area with warm lamp and plant, brown sports set. Text: “‘You don’t need discipline, you need better habits’”. Views: “31.9K”.
11. Bottom-center: bathroom mirror selfie, black sports bra and leggings, glass shower and towels behind her. Text: “What I eat when I’m craving everything”. Views: “74.1K”.
12. Bottom-right: mirror selfie in bright room with white doors and plant, cream sports bra and brown leggings. Text: “It’s not about eating less, it’s about eating right”. Views: “89K”.

Text styling: All overlay captions are centered near the top of each tile in clean bold white sans-serif text, Instagram-style, with subtle shadow for readability. Use {argument name="content theme" default="healthy eating and staying lean without obsessing over calories"} for the message tone. Keep all visible text in English and match the quoted phrases exactly.

Constraints: Make it look like a single cohesive Instagram Reels grid screenshot, not a collage poster. Keep the same creator across all 12 tiles. Do not add extra tiles, extra captions, usernames, watermarks, app navigation bars, likes, comments, or unrelated text. Use {argument name="interior palette" default="warm beige, taupe, cream, black, olive, and soft brown"} and {argument name="camera feel" default="realistic handheld phone mirror-selfie photography"}.

Using the attached image as the single source of truth, create ONE image that is a clean 3-column by 2-row grid of six photographs of the EXACT same scene, captured from six different camera positions, as if six cameras were placed around the room at the same instant.

Lock everything from the attached image: the same people with the same faces, hair, skin tone, glasses, clothing and body proportions; the same furniture, props, walls, signage with the same text, lighting setup and color grade; the same time of day. Nothing is added, removed, or rearranged — only the camera position, lens and framing change between panels. Every object stays in the same physical place in the room, so the spatial relationships remain consistent from angle to angle.

Panel 1 (top-left): Wide master shot — the same framing and camera position as the attached image, eye-level, 35mm lens, the full scene visible.

Panel 2 (top-center): High-angle top-down shot — camera mounted directly above the center of the scene looking straight down at the floor, 24mm lens, showing the layout of the furniture, props and people from above, like a ceiling camera.

Panel 3 (top-right): Over-the-shoulder shot from behind the subject on the LEFT of the original image, looking across at the subject on the RIGHT, 50mm lens at f/2.0, the near shoulder soft in the foreground, the far subject sharp.

Panel 4 (bottom-left): Reverse over-the-shoulder shot from behind the subject on the RIGHT of the original image, looking across at the subject on the LEFT, 50mm lens at f/2.0, near shoulder soft, far subject sharp.

Panel 5 (bottom-center): Low-angle three-quarter shot — camera placed low near floor height at the front-left corner of the scene, tilted slightly upward, 28mm lens, showing the scene with depth toward the back wall.

Panel 6 (bottom-right): Tight close-up on the main subject's face and upper body, 85mm lens at f/1.8, eye-level, shallow depth of field, background softly blurred, natural skin texture, catchlights in the eyes.

If the scene contains only one person, apply the over-the-shoulder panels as a three-quarter side angle from the left and from the right of that person instead. If the scene contains no people, treat the main object or furniture group as the subject for panels 3, 4 and 6.

Grid rules: six equal-sized panels, thin clean white gutters between them, no borders, no text labels, no numbers, no captions, no watermarks inside the image. All six panels share identical photographic style: photorealistic, same white balance, same exposure, same color grade, same subtle film grain, consistent lighting direction as in the attached image. Sharp, professional, real-camera look in every panel.

Photorealistic editorial photograph of a real podcast interview inside a small, cozy home studio. Two people are seated and talking to each other on a podcast set.

The person from Image 1 sits on the LEFT in a curved light-grey fabric armchair, body angled about 30 degrees toward the right, one hand resting naturally on the armrest, relaxed posture, mid-conversation with a genuine easy smile, eyes on the other person. The person from Image 2 sits on the RIGHT in a matching curved light-grey fabric armchair, body angled about 30 degrees toward the left, leaning slightly forward with elbows on knees, listening attentively with a warm expression, eyes on the person from Image 1.

Keep the exact same faces, skin tone, hairstyle, facial hair, glasses if present, body build and age as in Image 1 and Image 2. Do not alter or idealize either face. Clothing: keep each person's real outfit from their photo, clean and simple.

Between them stands a low white round side table with a small potted green plant. In front of each person a black professional broadcast microphone on a matte-black articulated boom arm, positioned near their mouth but not covering the face, cables tidy.

Background: a warm off-white plaster wall. Centered above and behind the two chairs, a black rectangular wall sign with the podcast name "{argument name="podcast name" default="PODCAST NAME"}" in clean white bold sans-serif capitals. To the right of the sign, a framed black-and-white typographic poster leaning on a small dark wood shelf; a dark wooden shelving unit on the far right with a few books, a small camera and a decorative object. On the far left a tall green indoor plant and a slim floor lamp. Dark grey curtains hang softly on the far left edge of the wall. Light wood floor.

Lighting: one large white softbox key light on a stand in the upper-left corner, softly illuminating both people at about a 2:1 ratio with smooth falloff; a large black-backed softbox visible in the upper-right corner as fill; warm practical glow from the floor lamp. Soft shadows, gentle wraparound skin light, natural skin texture with pores and subtle specular highlights, no harsh hotspots.

Camera: shot on a full-frame digital cinema camera

Create a Cannes-level flagship advertising poster for an original premium granola brand named {argument name="brand name" default="AURAGRAIN"}, featuring the product {argument name="product name" default="Honey Cinnamon Oat Granola"}. Build the image with a refined Orbit composition and a cooler, more international editorial aesthetic.

Core composition:
A vertical luxury food campaign poster with a soft mineral-off-white background, subtle natural paper texture, very clean negative space, and a calm premium atmosphere. In the exact center, place a polished silver spoon vertically as the hero object, filled with ultra-realistic honey cinnamon granola. The spoon must dominate the composition and feel precise, sculptural, and appetizing.

Hero product logic:
Inside the spoon, show detailed toasted oat flakes, compact crunchy granola clusters, roasted almond pieces, and a restrained glossy honey-coated texture. The food must feel authentic, premium, dry-crisp yet lightly glazed, with real commercial food-photography detail. No mess, no excessive crumbs, no rustic chaos.

Orbit structure:
Arrange 4 key ingredient elements around the spoon in a controlled circular flow: oat flakes, granola clusters, roasted almonds, and a small drop of golden honey. Their placement should create a subtle closed-loop visual rhythm around the spoon, guiding the eye in an elegant orbit. Use very thin refined directional lines and understated curved graphic gestures to connect the surrounding ingredients back toward the center. The orbit should feel intentional, premium, and quiet, not playful or messy.

Typography system:
Use a highly refined international typography layout. Replace any overly casual rustic feeling with a more elevated editorial balance:

- Large elegant custom script headline near the upper center-left: “{argument name="headline" default="Awaken"}”
- Small uppercase sans-serif line beneath it: “CLEAN ENERGY. QUIET LUXURY.”
- Small refined flavor descriptor at the upper right: “honey cinnamon" 
- Minimal ingredient annotations placed near each ingredient in tiny sophisticated English editorial type:
  “slow-toasted oats”
  “granola crunch”
  “roasted almonds”
  “pure honey glaze”

Typography must feel art-directed, airy, and premium, with beautiful spacing and hierarchy. No heavy fonts, no cheap health-product design, no crowded copy.

Branding:
Place a minimal AURAGRAIN logo in the upper left corner, understated and official. Add a very small net weight line in the lower right corner. Branding should feel discreet, int

Take the person from the reference image and create a single 2x2 grid showing this exact same close-up portrait under four different professional lighting setups. Identity lock: identical face, expression (calm neutral), hairstyle, dark top, head size and camera angle in every panel — only the light changes. Panel 1 REMBRANDT: key at forty-five degrees left and above, small triangle of light on the shadowed cheek, moody falloff. Panel 2 BUTTERFLY: key directly above the camera, soft symmetric shadow under the nose, beauty-portrait glow. Panel 3 SPLIT: key exactly from the side, one half of the face lit, the other in darkness. Panel 4 RIM: key behind the subject edging hair and shoulders in bright outline, face lifted only by faint fill, near-silhouette drama. Same {argument name="background color" default="medium-gray"} background throughout, 85mm look, sharp focus on the eyes. Small white labels beneath each panel: "REMBRANDT", "BUTTERFLY", "SPLIT", "RIM".

A realistic digital portrait of a beautiful {argument name="ethnicity" default="East Asian woman"} as a {argument name="character" default="modern angel"} standing on a high rooftop overlooking a dense, sprawling city under a bright daytime sky. She has long, dark flowing hair decorated with an intricate silver tiara, striking blue-grey eyes, and gentle feature makeup She wears an elegant {argument name="dress" default="white high-neck long-sleeve dress"} adorned with detailed gold embroidery around the chest and waist. Massive, highly detailed white feathered wings curve softly around her body Soft white feathers float through the air around her High angle perspective, natural daylight illuminating the cityscape in the background, hyper-detailed textures photorealistic cinematic lighting, 8k resolution.

A cinematic, ultra-realistic editorial portrait of a {argument name="subject" default="handsome adult man"} with tousled dark wavy hair, thick neatly trimmed beard, strong jawline, and an intense confident gaze directly at the camera. He is wearing a {argument name="clothing" default="slightly loose white button-down shirt"} with the top several buttons undone, sleeves casually rolled up, tucked into tailored black trousers with a black leather belt. A luxury metal wristwatch is visible on his wrist, one hand resting inside his trouser pocket. Dramatic low-key studio lighting from one side creates strong chiaroscuro, illuminating half of his face and shirt while the other side falls into deep shadow. Pure black minimalist background, moody masculine atmosphere, sophisticated luxury fashion editorial aesthetic, high contrast, realistic skin texture, subtle fabric details, natural anatomy, shallow depth of field, {argument name="lens" default="85mm"} portrait lens, vertical full-body composition, photorealistic, cinematic color grading, extremely detailed, premium fashion magazine photography.

Using the attached photo as the identity reference, create a photorealistic, heartwarming portrait of this exact same person as a happy 1-year-old baby.

Carry the person's real features into the baby version so the resemblance is clear: the same skin tone, eye color and eye shape, the same eyebrow shape, nose shape, lip shape, face proportions, hair color and hair texture, plus any distinctive traits such as dimples, a cleft chin, freckles or a particular smile. The baby's face is soft, round and chubby-cheeked as a real 12-month-old, with fine wispy baby hair in the person's natural hair color, and big bright eyes.

Gender is inferred from the attached photo. If the person is male, the baby wears a soft cotton outfit in his palette — a cozy long-sleeve onesie or a small t-shirt with soft pants and tiny socks. If the person is female, the baby wears a soft cotton outfit in her palette — a cozy long-sleeve romper or a little top with soft leggings and tiny socks. All clothing is comfortable, opaque and fully covering the body, with real fabric texture and natural creases.

Action: the baby is sitting on a plush carpet or a soft cream blanket on the floor, actively playing — holding up a colorful wooden stacking ring toy in both small hands, or reaching for soft building blocks, or hugging a small plush teddy bear. Natural baby posture, slightly wobbly, one leg tucked, fully absorbed in play. The baby has just looked up at the camera with a wide, genuine, gummy smile showing a couple of tiny front teeth, eyes crinkled with joy, cheeks lifted.

Setting: a bright, clean, cozy family living room — warm wooden floor, soft cream blanket, a few colorful wooden toys scattered around, a sofa softly blurred in the background, a window on one side letting in daylight. Everything is tidy, safe and homely.

Camera and light: shot on a full-frame camera with an 85mm lens at f/1.8, camera at the baby's eye level close to the floor, tight three-quarter body framing with the toy in frame, shallow depth of field, baby's eyes tack sharp, background softly blurred. Lighting is soft natural window daylight from one side with gentle fill, warm and airy, natural baby skin texture with soft highlights, clear catchlights in the eyes.

Style: looks like a real professional family photograph taken at home, candid and joyful, with a light warm color grade, subtle film grain and a palette of 60% soft cream and warm wood tones, 30% gentle pastel colors from the outfit and toys, 10% bright accent color from the stacking rings. Pure, wholesome and cute.

generate a {argument name="character" default="2d necromancer with a scythe"} as an animation ready game asset on a transparent background from {argument name="angles" default="all 8 angles"}

Using the attached photo as the identity reference, create a photorealistic, cinematic editorial portrait of this exact same person standing on a modern underground subway platform at night, in the style of a high-end street fashion photograph. Keep the exact same face, facial structure, skin tone, eye color, eyebrows, nose, lips, hairstyle, hair color, facial hair, glasses if present, age and body build as in the attached photo, so the person is instantly recognizable. Only the clothing, setting and photographic style change. Wardrobe adapts to the person: if male, a long black wool overcoat over a black turtleneck or crew-neck sweater, dark tailored trousers, clean minimal look. If female, a long black wool overcoat over a black fitted turtleneck, dark straight trousers, minimal jewelry, hair styled naturally. All clothing is opaque, closed and fully covering, with real wool texture, soft creases and a slightly turned-up collar. Pose and expression: standing near the platform edge, body turned about 30 degrees away from the camera, head turned back toward the lens, one hand relaxed in the coat pocket, the other hanging naturally. Calm, confident, slightly serious expression, direct eye contact with the camera, lips relaxed, chin slightly lowered. Three-quarter body framing from mid-thigh up, the person positioned on the right third of the frame. Setting: a clean modern subway station with a long concrete platform, dark grey tiled floor with a yellow tactile safety strip along the edge, dark walls, cool white ceiling strip lights receding into the distance on the left. Behind the person a subway train is passing at speed, rendered as a long-exposure motion blur: horizontal streaks of silver-grey metal and warm-lit windows, with a bright electric-blue LED light strip along the train body smearing into a glowing horizontal line. Wet reflections on the platform floor pick up the blue and warm light. Camera and light: shot on a full-frame camera with a 35mm lens at f/2.8, slow shutter around 1/8 second so the train is blurred while the person remains perfectly sharp and still, eye-level camera, slight low angle. Lighting is low-key: cool white overhead station light as key from above, blue rim light from the train streak on one side of the face and coat, warm amber fill from platform lamps on the other side. Deep soft shadows, natural skin texture with pores, crisp catchlights in the eyes. Style: cinematic night editorial, high contrast with rich blacks, subtle film grain, palette of 60% deep charcoal and black, 30% cool steel-blue and electric-blue light, 10% warm amber accents. Looks like a real photograph from a fashion campaign shot on location in a metro station.

Ultra-realistic cinematic portrait of a {argument name="subject description" default="strikingly handsome young man in his early 20s"}, seated confidently on wide white marble steps, creating a powerful, mysterious fashion-editorial atmosphere. He has thick naturally wavy jet-black hair, slightly tousled, strong well-defined eyebrows, deep expressive dark-brown eyes, a straight sculpted nose, prominent cheekbones, a sharp masculine jawline, and subtle natural stubble. He wears a black ribbed beanie, sleek black rectangular sunglasses, an oversized {argument name="outfit" default="black tactical-style jacket"} over a fitted black shirt, relaxed black trousers, and premium black combat boots. His posture is calm and dominant, elbows resting naturally near his knees, with both hands gently holding the head of a {argument name="animal" default="majestic black-and-tan Doberman"} sitting alertly between his legs. Several anonymous black-clad security figures stand behind him on the staircase, softly blurred and partially obscured by atmospheric white-gray haze. Minimalist white architectural background, symmetrical composition, centered subject, strong depth, dramatic soft lighting, subtle shadows, luxury streetwear aesthetic, moody monochrome palette with natural skin tones, photorealistic skin and fabric textures, cinematic color grading, shallow depth of field, 85mm portrait lens, high detail, editorial photography, 8K, vertical composition, no text, no watermark.

Create a photorealistic 2×2 four-panel recipe collage showing the preparation of a creamy banana smoothie on a rustic dark wooden kitchen table.

Top-left: Several ripe yellow bananas, one peeled banana, a clear glass of milk, a small glass bowl of thick white yogurt, and a bowl of ice cubes neatly arranged as ingredients.

Top-right: Close-up of hands peeling a ripe banana and placing banana pieces into a transparent glass blender jar.

Bottom-left: Milk being poured from a transparent measuring cup into the blender containing banana pieces, yogurt, and ice cubes.

Bottom-right: Thick creamy banana smoothie being poured from the blender into a clear drinking glass, with another finished smoothie glass in the background and a few banana slices on the wooden table.

Warm natural kitchen lighting, realistic food textures, creamy pale-yellow smoothie, transparent glassware, shallow depth of field, rustic wooden background, clean composition, professional food photography, highly detailed, realistic hands, natural shadows, 4K photorealism, no text, no labels, no watermark.

Create an ultra photorealistic cinematic historical photograph in a 2:3 portrait aspect ratio, set inside a {argument name="location" default="European railway station"} during the {argument name="time period" default="1940s"}.

A mysterious {argument name="character" default="elegant woman"} stands alone on an old railway platform, waiting beside a vintage suitcase. She wears authentic 1940s clothing with a sophisticated tailored coat, gloves, classic hairstyle, and period appropriate accessories.

Steam rises from a vintage locomotive in the background while passengers move naturally through the station. Old clocks, railway signs, luggage, wooden benches, architectural details, and period accurate clothing should make the environment feel historically authentic.

Create a strong cinematic composition with the woman as the central focal point and the railway tracks leading deep into the background.

Moody overcast daylight mixed with soft station lighting, realistic steam, atmospheric depth, natural skin texture, authentic fabric details, subtle film grain, vintage photographic characteristics, cinematic shadows, restrained color grading, premium historical photography, extremely realistic, 8K detail.

No modern objects, no modern clothing, no futuristic elements, no cyberpunk, no neon lights, no text overlays, no logos, no watermark.

Aspect ratio: 2:3.

Subject: A {argument name="subject" default="young man"} sitting casually on the fingers of a {argument name="giant object" default="giant human hand"}, creating a surreal miniature-person illusion. He has curly brown hair, black sunglasses, light stubble, a white sweatshirt with small “DON’T QUIT.” text on the chest, loose navy blue trousers, and white and navy sneakers.

Style: Photorealistic surreal street photography, realistic miniature effect, natural proportions, subtle cinematic look, shallow depth of field, highly detailed skin and clothing textures.

Scene: The giant hand is positioned in the foreground, with the young man sitting naturally across the fingers as if they were a bench. He sits relaxed with one knee raised and one leg hanging down. The hand wears several simple silver rings and a vintage rectangular gold wristwatch is visible near the lower right corner.

Background: {argument name="background" default="Elegant city townhouse entrance"} with large dark blue double doors, pale stone walls, black iron railings, and stone steps. The background is softly blurred to emphasize the miniature subject and giant hand.

Camera: Eye-level perspective, close-up foreground composition, 35mm lens, realistic depth of field, sharp focus on the man and hand, soft background bokeh.

Lighting: Soft natural daylight, slightly overcast, balanced exposure, realistic shadows, warm skin tones.

Aspect ratio: 4:5 vertical.

Ultra-realistic cinematic travel photograph of {argument name="subject" default="a strikingly handsome young man in his early 20s"}, portrayed as a giant standing {argument name="location" default="beside the Statue of Liberty in New York Harbor"}, creating a dramatic miniature-scale illusion. He has thick naturally wavy jet-black hair styled slightly tousled, strong well-defined eyebrows, deep expressive hazel-brown eyes, a straight sculpted nose, prominent cheekbones, a sharp masculine jawline, and subtle natural stubble. He wears {argument name="outfit" default="a luxurious deep burgundy oversized fleece jacket over a fitted black T-shirt, relaxed light-blue wide-leg jeans, and premium tan leather boots"}.He casually leans one arm against the Statue of Liberty's upper body while standing confidently beside it, relaxed fashion-model posture, natural facial expression. Low-angle perspective from near ground level, making both the man and statue appear enormous. New York City skyline visible across the Hudson River, including One World Trade Center, boats and ferries moving through the water, tiny tourists and pedestrians below emphasizing the massive scale.Bright clear blue sky with soft dramatic clouds, warm late-afternoon sunlight, realistic reflections on the water, detailed Statue of Liberty texture, cinematic depth of field, realistic skin pores and fabric texture, natural shadows, high-end editorial travel photography, photorealistic, HDR, dynamic composition, 9:16 vertical frame, extremely detailed, realistic proportions, no text, no watermark, no distortion.

Using uploaded face as reference. 
Study the uploaded photo's facial features and use them as inspiration to draw a similar-looking {argument name="subject" default="young man"} in a textured oil painting portrait style, thick impasto brushstrokes.
Hair: Messy, tousled curly dark hair falling naturally (no cap), light stubble beard.
Outfit: {argument name="clothing" default="White shirt with black tie"}, casual relaxed fit, no jewelry, no accessories.
Skin: Clean bare skin, no tattoos.
Pose: Close-up upper body portrait, one hand near chin/jaw, confident intense gaze directly at camera, shoulders slightly angled.
Background: {argument name="background style" default="Bold colorful graffiti street-art background"} — vibrant spray-paint style lettering and tags in orange, pink, purple, yellow, and teal colors, urban wall texture, energetic layered graffiti words like "STAY WILD," "ONE LOVE," and abstract graffiti symbols/arrows scattered around.
Art style: Thick textured oil painting, heavy palette-knife brushstrokes, realistic detailed face rendering blended with rough painterly texture, vivid saturated colors, high contrast, gallery-style fine art portrait.
Camera angle: Eye-level, close-up chest-up portrait shot, centered composition.
Text overlay:
Small stylish signature-style text in the top corner: "AMIR"

A full-body portrait of {argument name="subject" default="a beautiful middle-aged woman"}. The background is {argument name="background" default="a small courtyard in a Suzhou garden"} with a lake and blooming crabapple flowers. The woman is wearing a {argument name="outfit" default="light green Qipao"}, holding a round fan, sitting on a bamboo chair under a crabapple tree with her head slightly tilted up. It captures a fresh post-rain feeling. She has a full figure and elegant posture, with one side of the Qipao revealing her legs. Professional photographer work, Sony 35mm F1.4, background blur.

Oriental Zen minimalist cover poster, bright feminine aesthetic, new Chinese style morning feel high-value cover. 9:16 vertical composition, the overall image is light, clear, clean, and bright, with a modern Oriental feminine aesthetic and a Little Red Book (Xiaohongshu) cover feel.

The main subject is a young adult Oriental ancient style woman, {argument name="pose" default="standing by a giant circular lattice window"}. The person is located in the bottom right area of the frame, slightly turned, with a relaxed, quiet, and soft posture, carrying the ease and reserved temperament of having just woken up in the morning. The woman's features are delicate and refined, Oriental aesthetic, with a simple and elegant hair bun decorated with a few small white flowers and lightly drooping hair ribbons. The clothing is a light ancient style long dress with a {argument name="outfit colors" default="gradient of light white, pale apricot pink, and very pale mint green"}. The fabric is light and transparent, not overly gorgeous, without complex embroidery or heavy headpieces. The left hand naturally rests on the window sill, and the right hand hangs down holding a small sprig of white flowers. The hand structure is natural and accurate in proportion.

The scene is a minimalist white-walled indoor space with cream-white walls as the high-brightness base. Large areas of blank space are reserved on the left side of the frame, suitable for title layout. The circular window is the primary visual memory point. Outside the window, layers of mint green and light green plants are visible, along with a soft morning mist air layer, faint distant courtyards or mountain silhouettes, and an apricot-orange circular morning sun. The morning light is far outside the window, not stuck to the wall. Tree shadows are cast on the walls and floor, enhancing the morning light atmosphere.

The light is bright and soft natural morning light, with clean edges and textures, a slight paper feel is okay, no heavy ink feel, no grey fog, no old paper grain, no darkening. The overall temperament should be gentle, relaxed, light, clear, bright, and clean.

If a layout feel is needed, a small amount of vertical Chinese text can be added to the blank area on the left, such as: "{argument name="text" default="Morning light is just right / Window shadows of flowers / Grant yourself a slow period of time"}", accompanied by a small red seal, but the overall text must be restrained and not excessive.

Candid Ins-style lifestyle photography, {argument name="character" default="a beautiful 18-year-old Asian girl with double braids"} taking a selfie with the front camera, one hand propping up her cheek, the other resting casually on the edge of the bed. The image has a natural out-of-focus blur, like a moment captured right after waking up. The background is {argument name="environment" default="a bedroom during the day, with a crumpled knit cardigan and plush toys piled on the bed"}, overall {argument name="color tone" default="creamy white tone"}, with a dreamcore-like ambiguous and hazy atmosphere.

[Input Items] Jacket Color: [NAVY / BORDEAUX / FOREST GREEN / CAMEL] (Mandatory selection) [Purpose] Using only the single full-body portrait attached to this instruction manual as the sole character reference, create one design-focused simple character three-view sheet for both sharing and video reference. Inherit only the face, body type, and hairstyle from the reference image. Do not inherit the costume, footwear, accessories, bags, small items, or costume color scheme. Replace them with the specific costume for this occasion. [Aspect Ratio] Fixed at 4:3. Do not change to other ratios. [Character Reference] Maintain the facial features, eyes, nose, mouth, outline, age, and skin tone of the reference image. For the body type, maintain the proportions of the reference image, such as height, head-to-body ratio, shoulder width, torso length, waist position, hip position, leg length, and body slenderness/thickness. For the hairstyle, maintain the length, bangs, parting, hair volume, overall silhouette, and the shape of the sides and back from the reference image. Do not change the face, body type, or hairstyle due to clothing changes, and naturally develop as the same person in FRONT/SIDE/BACK views. [Costume Concept] Combine a neat and sophisticated tailored style for the upper body with a relaxed pajama style for the lower body. The upper body should give a sharp and high-quality impression, convincing as a sophisticated professional woman. Clarify the difference in texture and silhouette between the upper and lower halves to establish it as a single character design. [Jacket] A sophisticated tailored jacket. The design is fixed: a single-breasted jacket that naturally follows the body in a slightly slim fit, with one button, standard notched lapels, a moderately short length reaching near the hip bone, a naturally cinched waistline, and clean long sleeves. The shoulders should have a moderately structured and natural silhouette. Use only one of the colors selected in the [Input Items] (NAVY/BORDEAUX/FOREST GREEN/CAMEL) for the jacket body, maintaining the design even if the color is changed. [Inner] A high-quality minimal no-collar blouse fixed in OFF WHITE. The neckline should be a shallow and neat split neck, with a short vertical slit in the center of the neck. Do not open the chest widely like a general V-neck; keep the opening modest and sharp up to near the collarbone. Do not include lapels, shirt collars, collar tips, or stand-up collars. Do not create a collar like a general dress shirt. Do not use visible buttons on the front or a normal shirt placket. The material should be high-quality, smooth, matte blouse fabric, neither too thin nor transparent, and should not sag loosely on the body. It should not be excessively tight, but have a neat and sophisticated silhouette that naturally follows the body. Ensure the shape of the split neck is clearly and beautifully visible from the V-zone of the jacket, preventing collars or excess fabric from sticking out. Keep the neckline clean and organized to give a professional, intelligent, and urban impression. Maintain the same neckline, texture, and silhouette in FRONT/SIDE/BACK views. [Pajamas] Fixed PALE BLUE-GRAY long pajama pants for the lower body. Made of soft, light cotton-based material with a naturally relaxed straight silhouette from the waist to the hem. The waist should have a simple elastic specification. The fabric should have thin WHITE vertical stripes evenly placed, creating a clear loungewear design recognizable as pajamas. The color, pattern, silhouette, and length are fixed. [Footwear] Simple indoor slippers fixed in LIGHT GRAY. Maintain the same shape and color in FRONT/SIDE/BACK views. [Accessories] Do not use accessories. [Three-View Drawing] Left = FRONT / Center = SIDE / Right = BACK. FRONT is directly facing forward. Face, chest, waist, knees, and toes should point forward in a natural upright posture with both arms down naturally at the sides. SIDE is a complete profile, fixed facing left. Face, tip of the nose, chin, gaze, chest, abdomen, knees, and toes should point toward the left of the screen without twisting the body. BACK is a complete rear view. A natural upright posture where the back of the head, shoulders, back, waist, legs, and heels can be confirmed from behind. All three views must show the figure completely from the top of the head to the bottom of the slippers, unifying character size, height, head-to-body ratio, shoulder width, waist position, and leg length. The jacket, split-neck no-collar blouse, pajamas, and indoor slippers must also be established as the same design in all three views. [Fixed Layout] The 4:3 canvas is fixed into four areas from the left: "DESIGN STRIP | FRONT | SIDE | BACK". The leftmost 8% = DESIGN STRIP, and the remaining 92% = the three character regions for FRONT/SIDE/BACK, with all three character regions having the exact same width. The character regions are the main focus, while the DESIGN STRIP serves only as a visual accent for the sharing sheet. Characters are placed independently within each region at the same display magnification. [DESIGN STRIP] In the thin vertical area on the left edge, place only the rotated English heading "DESIGN" and four color swatches. "DESIGN" should be a normal horizontal English word rotated 90 degrees clockwise and placed vertically. The letters themselves should not be deformed; use a bold to medium-bold condensed sans-serif in all caps. Place it in the center of the DESIGN STRIP horizontally, starting at about 5-6% from the top edge. [Color Swatches] The four swatches are fixed from top to bottom in the order: NAVY / OFF WHITE / PALE BLUE / LIGHT GRAY. All should be squares of the same size, arranged in a single vertical column at the bottom of the DESIGN STRIP. Each side of a swatch should be about 50% of the DESIGN STRIP width, horizontally centered. The starting position of the swatch block is about 58% from the top edge of the canvas, with even spacing of about 2% of the canvas height between each swatch. Even if the jacket color is changed in the [Input Items], maintain the swatches as a fixed graphic element for the sheet design. [Character Headings] Display FRONT / SIDE / BACK exactly once at the top of each character region. The three headings must have the same font, weight, size, height, and top margin, using a condensed sans-serif from the same family as "DESIGN". FRONT / SIDE / BACK should be written horizontally. [Background / Separators] The background is white to a very light neutral gray. Use only thin outer border lines and thin vertical lines separating DESIGN STRIP / FRONT / SIDE / BACK. [Completion Conditions] 4:3 aspect ratio. A thin DESIGN STRIP on the left, and FRONT/SIDE/BACK in equal widths on the right. The face, body type, and hairstyle inherited from the reference image must be consistent as the same person in all three views. FRONT must be directly forward-facing, SIDE must be a complete profile facing left, and BACK must be a complete rear view. Unify character size, proportions, height, shoulder width, waist position, and leg length across the three views. The upper body must consist of the fixed tailored jacket + fixed split-neck no-collar blouse, the lower body must be the fixed pajamas, and the footwear must be the fixed slippers. Do not create a general shirt collar at the neck of the inner garment. The shallow split neck should look neat from the V-zone of the jacket. The neckline should not be messy with collars, fabric, or plackets, and the upper body should establish the character as a sophisticated professional woman. Use the color selected in the [Input Items] for the jacket, and maintain the position, structure, and order of the DESIGN STRIP and the four swatches. [Prohibitions] Aspect ratios other than 4:3, turning into a different person, changing the face/body type/hairstyle of the reference image, inheriting costume elements from the reference image, making the SIDE view face right or 3/4, cutting off parts of the body, overlapping characters, inconsistent character magnification between the three views, and changing the fixed layout or the four swatches. Do not add general shirt collars, fold-over collars, collar tips, stand-up collars, conspicuous plackets, or visible front buttons to the inner garment. Do not change the split neck to a deep V-neck. Do not add decorations such as frills, ribbons, bow ties, necklaces, or pendants to the neck.