DigiAI
پرامپت واقعی، نتیجه‌ی واقعی

هر تصویر خفن،
یه پرامپت داره.

پرامپت دقیق تصویرهای ساخته‌شده با Nano Banana، Midjourney، Flux و ChatGPT Image رو ببین، کپی کن و همون نتیجه رو خودت بساز.

An edgy conceptual portrait for an alternative fashion label. A headshot captures a subject with dar
پرامپتGPT Image

@leylaمشاهده کامل
Create an ultra-realistic  using the reference image as the identity source.
Preserve the exact faci
Nano Banana

Create an ultra-realistic using the reference image as the identity source. Preserve the exact facial features, bone structure, proportions, skin tone, and identity of the subject. Do not beautify, reshape, or alter the face in any way. Maintain natural likeness and authentic facial geometry. Core Subject & Pose: Intimate black and white close-up portrait of a couple in a deep, emotional embrace. Female subject with her head resting gracefully against the man. The male subject has his face pressed tenderly against her temple. Both have eyes gently closed in a comforting, somber stance. Wardrobe & Styling: Solid black, heavy-textured garments blending seamlessly into the dark foreground shadows. Exact Makeup & Hairstyle: Female: Raw, natural makeup, prominent authentic freckles across the nose and cheeks, soft matte lips. Dark, heavily textured hair dramatically windblown upwards and to the left. Male: Wind-swept dark hair and a rugged short beard. Setting & Environmental Details: Muted, abstract stormy gray background suggesting a tempestuous, windy environment. Lighting & Expression Mapping: Moody cinematic chiaroscuro. Soft, directional spotlighting on the center of their faces, fading into pure black on the edges. Expressions map to deep emotional intimacy, peace, and protection. Camera Details & Optical: Shot on Phase One XF and Sony A7R V, 85mm portrait lens, f/1.8 aperture for shallow depth of field, razor-sharp optical focus on facial features. Composition & Angles: Tight framing, center-weighted, intimate eye-level angle focusing entirely on the intertwined faces. Realism & Texture Boosters: Ultra-realistic, ultra-HD, 8K resolution, true monochrome photography, fine skin pores, rendering of individual hair strands, micro-contrast on freckles and stubble, cinematic realism, authentic subtle film grain. Negative Prompt: color, vibrant, saturated, sepia, nose piercing, altered facial features, heavy makeup, airbrushed skin, plastic, 3D render, CGI, cartoon, anime, smiling, laughing, bright lighting, overexposed, cluttered background, props, extra limbs, distorted anatomy, poorly drawn face, deformed eyes, morphed features, watermark, text, signature, low resolution, blurry, grainy background, noisy, flat lighting.

پرامپت عکس تبلیغاتی:👇

نام محصول = [" نام محصولتون رو اینجا وارد کنید "]

"هدف": "ایجاد یک تبلیغ شب
GPT Image

پرامپت عکس تبلیغاتی:👇 نام محصول = [" نام محصولتون رو اینجا وارد کنید "] "هدف": "ایجاد یک تبلیغ شبکه‌ای 2×2 اصلاح‌شده برای محصول لوکس خانگی '[ نام محصول ]'، با تاکید بر حس، حال و هوا و ظرافت حسی از طریق تصاویر مینیمال و به سبک تحریریه."، "بوم نقاشی": "نسبت_ابعاد": "1:1"، "ساختار_شبکه": "پانل‌های مساوی 2×2"، "جهت": "مربع"، "مورد_استفاده": "رسانه‌های اجتماعی برند لوکس"، "کمپین‌های خانه و سبک زندگی"، "داستان‌سرایی محصول تجارت الکترونیک"، "تبلیغات دیجیتال به سبک تحریریه" "پانل‌ها": "شماره_پانل": 1، "عنوان": "شیء محصول قهرمان"، "توضیحات": "عکس قهرمان استودیویی از [نام محصول] که روی یک سطح طبیعی متمرکز شده است."، "جزئیات_بصری": "نور": "نور پراکنده ملایم با سایه‌های ملایم"، "زمینه": "پارچه یا سطح با رنگ خنثی [پالت رنگ]"، "تمرکز": "شکل محصول، کیفیت مواد، برچسب مینیمالیستی"، "حال و هوا": ["ظرافت"، "آرامش"، "پریمیوم"] "شماره پنل": 2، "عنوان": "فضای داخلی با حال و هوای دلنشین"، "توضیحات": "صحنه داخلی سبک زندگی با [نام محصول] قرار داده شده روی قفسه، میز یا سطح."، "جزئیات بصری": "محیط": "فضای زندگی مدرن و آرام"، "نورپردازی": "نور طبیعی پنجره"، "حال و هوا": ["راحتی"، ​​"گرما"، "زندگی آرام"] "شماره پنل": 3، "عنوان": "جزئیات آیینی"، "توضیحات": "نمای نزدیک از دست‌هایی که با [محصول] در تعامل هستند نام] در ژست کاربرد اصلی آن."، "جزئیات_بصری": "تمرکز": "تعامل انسانی و آیین"، "عمق_میدان": "کم‌عمق و صمیمی"، "حال و هوا": ["حسی"، "شخصی"، "ذهن آگاه"] "شماره_پنل": 4، "عنوان": "جنس و حس و حال"، "توضیحات": "ترکیب‌بندی ماکرو یا طبیعت بی‌جان که محصول را در کنار [مواد حسی] نشان می‌دهد."، "جزئیات_بصری": "عناصر": "[مواد حسی]"، "نورپردازی": "هایلایت‌های ملایم، کنتراست کم"، "حس و حال": ["طبیعی"، "بی‌زمان"، "فضایی"] "برندسازی": "نام_محصول": "[نام محصول]"، "شخصیت_برند": "لوکس آرام، مدرن، متمرکز بر حس"، "پالت_رنگ": "[پالت_رنگ]"، "ارائه_لوگو": "مینیمال، کم‌رنگ، متمرکز یا برجسته" "سبک_بصری": "کلمات_کلیدی_سبک": "تبلیغات سبک زندگی خانه‌های لوکس"، "عکاسی طبیعت بی‌جان سرمقاله‌ای"، "ترکیب‌بندی مینیمالیستی"، "واقع‌گرایی بالا"، "رنگ‌های طبیعی ملایم" "دوربین و نورپردازی": "نوع_عکاسی": "عکاسی ترکیبی سرمقاله‌ای و تجاری"، "انواع_عکس": "عکس محصول قهرمان"، "عکس سبک زندگی داخلی"، "جزئیات ماکرو"، "ترکیب‌بندی طبیعت بی‌جان" "سبک_نورپردازی": "نور طبیعی پنجره با پخش ملایم"، "نوردهی": "کنتراست کم، گرم، متعادل"، "سایه‌ها": "نرم و ارگانیک" "کلمات_کلیدی_حال_و_هوا": "آرام"، "تصفیه‌شده"، "فضای_دوستانه"، "زیبا"، "گرم"، "حسی" "موارد مممنوعه و منفی": «نورپردازی شدید»، «فضای داخلی نامرتب»، «ظاهر متریال ارزان»، «رنگ‌های پررنگ یا جیغ»، «برچسب‌های ناخوانا»، «ظاهر CGI یا مصنوعی»، «طراحی بیش از حد»، «واترمارک»، «پوشش متن»

pencereden gelen güçlü dogal gün sigina benziyor.
• GÖrselin genelinde güçlü gölgeler görünüyor.
• G
GPT Image

pencereden gelen güçlü dogal gün sigina benziyor. • GÖrselin genelinde güçlü gölgeler görünüyor. • Genel renk paleti krem, bej, yumusak pembe ve sicak kahve tonlarindan oluguyor Üç Kareli Kompozisyon Üst Kare: Yüzün yarisi görünüyor, bir göz ve dudaklar odak noktasi. Kamera bu kareyi egik açiyla yakaliyor. Orta Kare: En vakin cekim, gözün net odaginda. iris detaylari, cilt dokusu gözenekler ve kirpikler keskin sekilde görünüyor. 3. Alt Kare: Biraz daha genis portre, iki qözü, dudaklari ve alnin üzerine düsen gevsek sac tellerini gösteriyor, daha duygusal ve sinematik bir his yaratiyor. Genel olarak görsel dogal, vumusak feminen ve hafif sinematik portre estetigi yansitmali. Ana odak parlak gözler ve dolgun dudaklar üzerinde Teknik Gereksinimler: • Ultra gerçekçi fotograf • Dikey 9:16 en-boy orani• Güclü dogal isik, gözler ve yüzün bir kismina odakl • Görünür gözeneklerle gerçekçi cilt dokusu • Yüksek detayli sac telleri 4K kalite . iPhone 16 Pro Max ile cekilmis • Plastik görünümlü cilt vok • Yapay güzellik filtrelerinin izi yok • Dogal vüz ifadesi ve vücut durusu Referans yüz degistirilmeden ayni ka.copy my face 100%

Restore and colorize this historical photograph while preserving 100% of the original composition, c
GPT Image

Restore and colorize this historical photograph while preserving 100% of the original composition, camera angle, perspective, architecture, people, vehicles, signs, and every authentic detail. Do not add, remove, replace, or alter any objects. Perform museum-grade photo restoration by removing scratches, dust, stains, tears, film grain, noise, blur, fading, and compression artifacts while maintaining the original texture. Apply historically accurate colorization using realistic skin tones, clothing colors, architectural materials, vegetation, road surfaces, vehicles, and environmental lighting appropriate for the era and location. Avoid oversaturation or artificial colors. Enhance fine details including faces, hair, fabrics, bricks, wood, metal, glass, stone, shadows, reflections, and depth while preserving authenticity. Correct exposure, white balance, dynamic range, contrast, and lighting naturally. The final result should look like a genuine modern high-end DSLR photograph captured at the same moment in history. Ultra realistic, museum-quality restoration, 8K resolution, ultra sharp, photorealistic, cinematic natural lighting, maximum detail, HDR, true-to-life colors, no AI artifacts, no hallucinations, no added objects, no text changes.

An edgy conceptual portrait for an alternative fashion label. A headshot captures a subject with dar
GPT Image

An edgy conceptual portrait for an alternative fashion label. A headshot captures a subject with dark features beneath a heavy brown leather cowboy hat, accessorized with stacked silver jewelry and a patterned neck bandana. Intense side-lighting models her facial structure against a warm neutral backdrop, creating a moody, cinematic film still quality.

Строго сохранить внешность 1:1 по загруженным фото: черты лица, пропорции, возраст, форму глаз, носа
Nano Banana

Строго сохранить внешность 1:1 по загруженным фото: черты лица, пропорции, возраст, форму глаз, носа, губ, естественную асимметрию лица; без идеализации, морфинга, face swap и beautify. Формат 9:16. Фотореалистичный студийный семейный портрет в минималистичном стиле. Черный бесшовный фон. Мама сидит ближе к камере, опираясь подбородком на обе ладони, локти слегка сведены, спокойная нежная улыбка и прямой взгляд в объектив. Сын стоит позади мамы, наклонившись вперед, обе ладони лежат на макушке мамы, подбородок опирается на руки. Мальчик улыбается, взгляд направлен прямо в камеру. Между ними чувствуется искренняя теплота и легкая игровая атмосфера. На маме легкое платье с мелким цветочным принтом, на сыне простая светлая футболка без принтов. Макияж у мамы: сияющая кожа, легкий контуринг, естественный румянец, нюдовые губы, длинные ресницы, подчеркнутые глаза. Волосы объемно уложены. Мягкий студийный свет через большой октабокс, плавные объемные тени, глубокий черный фон, классическая портретная обработка в черно-белой гамме с мягким контрастом, высокой детализацией кожи и волос, атмосферная fine art photography. 85 mm, f/2.0, shallow depth of field, photorealistic skin texture, black and white portrait, fine art studio photography, ultra realistic, high detail, 8K.

What would I have looked like if I lived in Iran during the 1980s, styled in the typical fashion of
Nano Banana

What would I have looked like if I lived in Iran during the 1980s, styled in the typical fashion of that era, with natural uncovered hair (no hijab or chador)? I am standing in a traditional Iranian setting such as old Tehran streets or a vintage Persian home.The scene should be next to a sandwich shop, where we are sitting or standing together, smiling and eating sandwiches in a lively and friendly atmosphere.Please style the hair according to the 1960s or 1980s Iranian fashion.The environment must not include any wall photos, portraits, or visible images of people in the background.Make sure the faces remain unchanged and identical to the original images.

MAINTAIN 100% similarity: do not change facial features, proportions, shape of nose/lips/eyes, hair
Nano Banana

MAINTAIN 100% similarity: do not change facial features, proportions, shape of nose/lips/eyes, hair color and texture. The face is recognizable.Black and white photograph. Spontaneous night shot, a moment accidentally caught in motion. The frame looks like a real amateur photo, not a staged fashion shoot.POSE: the girl is sitting on a transparent acrylic chair, torso in 3/4 and slightly leaning forward, both hands resting on her knees. Head tilted slightly down, but at the moment of the shot she sharply turns it to the side. In her hands is a large, voluminous bouquet of light-colored flowers, which partially covers the lower part of the frame. Gaze is averted, not directly into the lens. Head movement creates natural motion blur on the face and hair, while the face remains recognizable.HAIR: casually tied up, strands near the face blurred by movement.CLOTHING: white oversized t-shirt, casual.BOUQUET: large, voluminous bouquet of light-colored flowers, arranged naturally and slightly casually. Flowers are close to the body and partially cover the lower part of the frame. Individual flowers and petals are slightly displaced by movement, creating a lively texture. The bouquet is the main visual anchor of the scene.ENVIRONMENT: minimalist space with a neutral gray background, no extra decor. The girl and the bouquet are visually separated from the background by natural soft light. No sterile studio perfection.BLACK AND WHITE PROCESSING: strict monochrome, deep blacks, silver-gray midtones, and soft milky highlights. Natural analog contrast, no color tints.Vintage film: noticeable analog grain, slightly faded film texture, light vignette, soft contours, slight focus falloff. The image fills the entire frame from edge to edge, NO white border, full-bleed. Handheld, slightly imperfect framing. Raw unedited frame without retouching.Do not use 8K, ultra sharp, high-detail enhancement, no blur, HDR, oversaturation, plastic skin, beauty retouching, or sterile digital sharpness.--ar 9:16 --raw

Transform the people in the uploaded photo into photorealistic LEGO minifigures. Use the uploaded ph
Nano Banana

Transform the people in the uploaded photo into photorealistic LEGO minifigures. Use the uploaded photo as the only reference for their identities, hairstyles, hair colors, facial hair, expressions, outfits, and wearable accessories. Preserve recognizable characteristics as closely as LEGO styling allows. Keep the same number of people, their poses, and their relative positions. Use authentic LEGO proportions: cylindrical yellow heads, simplified printed facial features, blocky bodies, articulated limbs, and classic C-shaped hands. Render smooth, glossy yellow ABS plastic with realistic reflections. Recreate the original clothing colors and designs as clean toy details, removing any lettering or branding. Replace the original background with a seamless pure-white studio background. Show only the LEGO figures, centered and fully visible, with balanced white space around them. Soft, bright studio lighting, gentle contact shadows beneath their feet, crisp details, premium 3D toy product photography. Vertical 4:5 composition. No text, letters, numbers, logos, watermarks, titles, captions, labels, badges, slide counters, frames, borders, inset photographs, Polaroids, packaging, or graphic overlays. Do not add people, props, or decorative elements.

a young gym boy standing in the forest and ice falling around and the weather is amazing
GPT Image

a young gym boy standing in the forest and ice falling around and the weather is amazing

Combine the best natural expression of each person from the uploaded photos into ONE photorealistic
Nano Banana

Combine the best natural expression of each person from the uploaded photos into ONE photorealistic group photograph. Use one image as the base, keeping its composition, positions, poses, clothing, background, and lighting. If I have not selected a base, choose the sharpest, best-framed photo where everyone is sufficiently visible. Match each person carefully across the photos using facial features, clothing, and position; never assume their position stays the same. If identity matching is uncertain, ask before editing. For each person, prefer naturally open eyes, a relaxed expression, and clear focus. Preserve intentional closed eyes during laughter or affection. Do not force everyone to smile or look at the camera. Keep anyone who already looks good unchanged. Use only expressions and facial details actually visible in that same person’s reference photos. Never invent teeth, eye color, a smile, or hidden features. If no suitable alternative exists, retain the original. Make the smallest possible edits. Transfer only the necessary eye or mouth region when head angles match. If a larger facial region is needed, use a compatible reference and preserve believable head, jaw, neck, and hair connections. Do not force an incompatible angle. Match perspective, facial scale, skin tone, light direction, shadows, sharpness, depth of field, and grain. Preserve each person’s identity, age, facial proportions, skin texture, makeup, hair, and facial hair. Keep natural asymmetry and fine expression lines. Maintain glasses, accessories, overlapping hair, hands, and contact between people. No face blending, swapped identities, duplicated people, altered bodies, beauty filters, skin smoothing, or unrelated retouching. Use photos from the same scene or sequence whenever possible. If the uploads show incompatible scenes or lack usable references, briefly explain what additional photo is needed instead of fabricating a convincing match. The result should look like one authentic moment when everyone happened to look their best—not a collection of pasted faces. Return ONE finished photograph, with no comparison panels, captions, logos, or watermark.

Use the uploaded image as a STRICT VISUAL REFERENCE and recreate the exact photo.

IDENTITY — ABSOLU
Nano Banana

Use the uploaded image as a STRICT VISUAL REFERENCE and recreate the exact photo. IDENTITY — ABSOLUTE PRIORITY: Preserve both faces and identities exactly: facial structure, proportions, eyes, nose, lips, jawline, natural asymmetry, skin tone and texture. Do not beautify, redesign, copy-paste, blend or replace faces. HAIR — STRICT REFERENCE ONLY: The woman's hair MUST remain exactly as shown in her uploaded identity reference. Preserve the EXACT original hair color, shade, undertone, highlights, roots, darkness/brightness, and natural color variations from the reference. DO NOT change, recolor, lighten, darken, warm, cool, tint, or reinterpret her hair color under any circumstances. Also preserve her exact hairline, texture and density from the reference. The reference image is the ONLY source for her hair color. Woman taller, man broad-built. Woman stands before the mirror, body slightly angled, one leg forward; one hand holds a black triple-camera iPhone beside her face, the other holds a black structured handbag. Keep the woman's hair color EXACTLY from her uploaded reference, while styling it long and straight with a side part and side-swept bangs. Pink-red lipstick, short red acrylic nails. She wears a fitted black lace long-sleeve top with sheer sleeves, chest cutout, corset bodice and long black skirt. Man wears a black suit and turtleneck. Bright room, wooden floor, large arched wood-framed mirror. Authentic casual iPhone photo, natural pores, subtle grain/noise, slight lens softness, realistic shadows and exposure; no HDR, beauty filter, plastic skin, CGI or AI look.

پرامپت کاپلی
Nano Banana

پرامپت کاپلی

Use my childhood photo as the only identity reference. Create an ultra-realistic vertical 2:3 photo
Nano Banana

Use my childhood photo as the only identity reference. Create an ultra-realistic vertical 2:3 photo of me as a tiny toddler on a vast foggy gray road, body sideways and looking back calmly. Keep my exact face, skin tone, and dark curly hair in two pigtails. Orange oversized puffer jacket, cream trousers, beige sneakers, woven basket bag, large empty space above, soft overcast light, subtle film grain. No face alteration, straight bangs, text, or watermark.

Create a high-quality, photorealistic vertical beauty portrait using the uploaded photos as the exac
Nano Banana

Create a high-quality, photorealistic vertical beauty portrait using the uploaded photos as the exact identity reference. The woman must remain exactly the same person: preserve her face shape, eyebrows, eyes, nose, lips, teeth, smile, skin tone, and all facial proportions. Do not alter, beautify, reshape, or make her resemble anyone else. Create a close-up three-quarter portrait, with her eyes gently closed and a natural wide smile. Her long warm light-brown hair is loose, smooth, and softly voluminous. Use a dark charcoal studio background with a strong warm editorial spotlight illuminating one half of her face and hair, while the other half has soft cinematic shadow. Add a small artistic scattering of colorful rhinestone flower stickers and tiny round gems across the forehead, cheeks, and jawline in fuchsia, turquoise, violet, lime, and pink. Keep them realistic, reflective, and naturally attached to the skin. Use glossy vivid red lipstick and a satin fuchsia-pink top. Premium beauty-editorial photography, vertical 4:5 Instagram portrait, realistic skin texture, individual hair strands, sharp high-resolution quality, professional studio lighting. No text, no logo, no watermark, no plastic skin, no blurred details, no distorted eyes, no extra fingers.

Transform the uploaded photograph into a brush-painted, semi-realistic character illustration with a
GPT Image

Transform the uploaded photograph into a brush-painted, semi-realistic character illustration with a refined concept-art finish. REFERENCE PRESERVATION: Use the source photo as the sole reference for identity and composition. Preserve recognizable facial features, facial proportions, age, expression, gaze, hairstyle, pose, hands, props, clothing, camera angle, framing, and aspect ratio. Retain the original number of people and objects. PAINTING STYLE: Use visible directional brushstrokes and gently simplified facial planes. Describe the jaw and brow through painted light and shadow without reshaping them. Render hair as sculpted, flowing masses that follow the original hairline, volume, and silhouette rather than individual strands. BRUSHWORK AND TEXTURE: Use softly blended but visible strokes on skin, crisp palette-knife accents along fabric folds, and loose, broader strokes in the background. Keep the face and important props clearly resolved. Add subtle canvas grain and dry-brush texture along selected edges; avoid a uniform texture overlay. COLOR AND LIGHT: Use muted, rich colors inspired by peach #E8A88C, brick red #C4443A, slate blue #3E5C8A, dark brown #2A2420, and warm beige #D9CBB8. Treat these as grading accents, preserving the subject’s underlying skin tone and recognizable object colors. Create warm skin highlights with restrained flushed cheeks against a cooler background. Preserve the source lighting direction and coherent shadows. CONSTRAINTS: The result must read as an intentional painting, not a photograph with a filter. Avoid photographic pores, plastic or excessively airbrushed skin, smooth vector shading, exaggerated facial anatomy, distorted hands, missing props, added text, logos, orwatermarks.

Analyze the uploaded photo and create a feature-by-feature styling and makeup placement guide.
The f
Nano Banana

Analyze the uploaded photo and create a feature-by-feature styling and makeup placement guide. The final results must be as photo For each feature, assess its shape and proportion, then give specific guidance: Eyebrows: Current shape assessment, ideal arch position, recommended thickness, and where my brow should start, peak, and end relative to my nose and eye. Eyes: Eye shape (almond, round, hooded, monolid, downturned, etc.), spacing (close-set, wide-set, average), and specific eyeshadow placement to enhance my natural eye shape. Lips: Lip proportion (full, thin, uneven), cupid's bow definition, and lipstick or liner techniques to enhance shape. Cheekbones and Contour: Cheekbone placement and prominence, exact contour placement lines, where to place blush and highlighter for my specific bone structure. Nose: Nose shape and any contouring techniques that complement it. Be specific about product placement using anatomical landmarks. Honest assessment, not flattering. All the texts in photo should be persian/farsi iranian language

بدون تغییر ویژگی‌های صورت و مطابق چهره های مرجع آپلود شده، یک عکس پرتره سینمایی خلق کنید. چهره ها و
Nano Banana

بدون تغییر ویژگی‌های صورت و مطابق چهره های مرجع آپلود شده، یک عکس پرتره سینمایی خلق کنید. چهره ها و جزییاتش دقیقا به صورت 1:1 مطابق عکس های آپلود شده باشد. ساختار دقیق صورت، تناسبات و عدم تقارن طبیعی را حفظ کنید. یک پرتره بسیار نزدیک. یک کراپ بسته: فقط بخشی از صورت مرد (نیمی از صورت) قابل مشاهده است. یک زن مستقیماً پشت سر او قرار گرفته است؛ صورتش تا حدی پشت سر/شانه مرد پنهان است. زن در فاصله ی نزدیک به مرد است گویا او را از پشت در آغوش گرفته است. ترکیب‌بندی چندلایه قوی (مرد در پیش‌زمینه، زن در پس‌زمینه). مرد رو به جلو، با کت مشکی و پیراهن مشکی شیک. زن بسیار نزدیک پشت سر او با لباسی سفید ایستاده است. مرد: آرام، خنثی، حالت چهره کمی جدی. زن: نگاه خیره، لب‌هایی که به آرامی به هم فشرده یا کمی باز شده‌اند، آرایش بی‌عیب و نقص، موهای موج‌دار که به آرامی توسط باد تکان می‌خورد. نورپردازی و رنگ سینمایی و سرد نزدیک به سیاه و سفید، با بک گراندی با تاری زیاد و به رنگ تیره،

Use the uploaded photo(s) of the couple as the strict identity reference. Preserve both people's rec
Nano Banana

Use the uploaded photo(s) of the couple as the strict identity reference. Preserve both people's recognizable faces, facial structure, skin tone, hair, age, body proportions and overall identity consistently across all outputs. Create THREE separate ultra-photorealistic full-body vertical photos of the SAME couple, each as an individual image — NOT a collage, NOT a triptych. SETTING FOR ALL 3 IMAGES: A quiet residential street in Iran at night, extremely realistic and naturally Iranian: ordinary apartment buildings on both sides, Persian street signs, overhead cables, warm amber streetlights, asphalt road and realistic Iranian cars parked along both sides with Iranian license plates. Deep navy night sky. The atmosphere should feel like a genuine spontaneous smartphone photo taken by a friend, not a professional photoshoot. Keep the location, lighting, clothing and couple identical across all three photos. WARDROBE: Both wear simple oversized black T-shirts. Male: loose black cargo pants and black-and-white high-top sneakers. Female: loose high-waisted blue jeans and matching black-and-white high-top sneakers. Natural casual styling, no luxury fashion look. IMAGE 1 — FOREHEAD KISS / HUG: The couple stands closely in the middle of the street. The woman wraps both arms around the man's waist and rests against his chest with her eyes softly closed and a subtle happy smile. The man embraces her with one arm around her body, gently holds the side/back of her head with the other hand and kisses her forehead. Tender, intimate, natural body language. IMAGE 2 — FACE-TO-FACE: Same exact street and appearance. They stand very close, bodies touching. The woman has one arm around his waist and tilts her face upward, smiling lovingly into his eyes. The man looks down at her with a soft smile, one hand casually in his pocket and the other gently touching her hair/head. Romantic but completely candid and believable. IMAGE 3 — PLAYFUL TWIRL: Same exact street and appearance. The man raises one arm above the woman and holds her hand while she spins underneath it. Her long hair moves naturally from the turn, she smiles and looks slightly downward, while he watches her with a happy relaxed expression. Capture the movement at the perfect spontaneous moment. VISUAL STYLE: Ultra-photorealistic nighttime smartphone photography, vertical 9:16 composition, full bodies visible from head to shoes, natural perspective around eye/chest height, realistic anatomy and height difference, authentic skin texture, subtle smartphone sharpening and noise, slight low-light softness, realistic shadows on the asphalt, warm streetlight highlights, natural dynamic range and believable depth. The couple must look like the real people from the uploaded reference images — do not beautify them into different people and do not change their facial features. Do not add text, captions, borders, filters, watermarks or UI elements. Output exactly 3 separate images.

Use the uploaded image as the ONLY facial identity reference. Preserve the person’s exact facial ide
GPT Image

Use the uploaded image as the ONLY facial identity reference. Preserve the person’s exact facial identity, facial structure, proportions, skin tone, age, natural features, and overall appearance. Do not change, beautify, feminize, masculinize, or reinterpret the face. If the reference person is male, generate a male subject; if the reference person is female, generate a female subject. Keep the person’s natural gender presentation exactly consistent with the reference image. Recreate the scene as a highly realistic cinematic smartphone photograph inside a dark luxury car at night. The subject is sitting in the front seat, photographed from a close-up slightly front-facing angle. The person is resting one side of their face gently on one hand, with a relaxed, confident, effortless pose. Match the overall composition, framing, lighting, atmosphere, and aesthetic of the reference image: dark moody interior, black clothing, subtle warm low-light illumination on the face, deep shadows, dark background, and visible dark-red leather car seats in the background. Add a stylish black baseball cap and narrow black oval sunglasses naturally fitted to the subject’s face. Keep the accessories minimal and realistic. The hairstyle should be adapted naturally to the subject’s actual gender and facial features while maintaining the same overall aesthetic. Do not copy the reference person's identity, facial features, or gender—only use the uploaded image to preserve the identity of the new subject. Ultra-realistic photography, natural skin texture, realistic pores, accurate facial proportions, authentic human anatomy, subtle cinematic grain, soft low-key lighting, realistic reflections on the sunglasses, realistic fabric and leather textures, premium editorial street-style aesthetic, candid smartphone photography, high detail. CRITICAL IDENTITY RULE: The face must remain unmistakably the same person as the uploaded reference image. Do not alter the eyes, eyebrows, nose, lips, jawline, cheekbones, face shape, skin tone, age, or distinctive facial characteristics. Vertical 4:5 composition, close-up portrait, realistic photography, no text, no watermark, no artificial-looking skin, no exaggerated beauty filter.