Analyze the single image I upload. Identify my exact face shape and hair type with precision.
Then generate a vertical 9:16 4K 3x3 grid of hairstyle reference images directly on my face.
Requirements: Analyze the uploaded image to detect: Face shape, Hair density, texture, curl pattern, and growth direction. Based on this analysis, recommend haircuts for three scenarios: Professional or workplace, University or college, School or teenage. Provide 4 haircut options per scenario, each with a short, specific explanation of why it suits my geometry and hair behavior. After the recommendations, generate a 3x3 grid of haircut images, each placed realistically on my face. Technical requirements:
Vertical 9:16 orientation, 4K quality, Realistic lighting, Natural hair flow and believable texture,
No fantasy looks or exaggerated Al styling.
Variation rules for the 3x3 grid: Mix short, medium, and slightly longer styles, Reflect the recommended haircuts, Maintain consistency with my real hair type and proportions. Output must be structured, clean, and easy to compare across scenarios. Your goal is accuracy, realism, and professional grooming quality.
Then generate a vertical 9:16 4K 3x3 grid of hairstyle reference images directly on my face.
Requirements: Analyze the uploaded image to detect: Face shape, Hair density, texture, curl pattern, and growth direction. Based on this analysis, recommend haircuts for three scenarios: Professional or workplace, University or college, School or teenage. Provide 4 haircut options per scenario, each with a short, specific explanation of why it suits my geometry and hair behavior. After the recommendations, generate a 3x3 grid of haircut images, each placed realistically on my face. Technical requirements:
Vertical 9:16 orientation, 4K quality, Realistic lighting, Natural hair flow and believable texture,
No fantasy looks or exaggerated Al styling.
Variation rules for the 3x3 grid: Mix short, medium, and slightly longer styles, Reflect the recommended haircuts, Maintain consistency with my real hair type and proportions. Output must be structured, clean, and easy to compare across scenarios. Your goal is accuracy, realism, and professional grooming quality.
❤3
✅Here how you can create them:
⭐️Image prompt -
1. Use the uploaded image of the man as the primary face and identity reference.
Preserve the same facial structure, skin tone, hairstyle, beard/stubble, glasses, and overall likeness with high facial accuracy. Do not alter identity.
Create an ultra-realistic photo of this same man taking a front-facing selfie at an active construction site.
Outfit: beige/tan work jacket over a grey t-shirt, realistic fabric texture, same fit and color tone as reference.
Environment: unfinished buildings with scaffolding, excavators, bulldozers, dirt ground, industrial construction setting.
A large metallic wrecking ball is suspended from a thick steel cable directly above the man’s head, clearly visible and centered above him, creating dramatic tension.
Composition: selfie perspective, chest-up framing, subject centered, wrecking ball aligned vertically above head.
Lighting: natural daylight, clear blue sky, realistic shadows, outdoor realism.
Style: hyper-realistic photography, DSLR / smartphone selfie realism, sharp focus, natural colors.
Strict rules:
– Maintain face consistency with the uploaded image
– No face reshaping
– No stylization or cartoon look
– No blur, no distortion, no AI artifacts
– No text, no watermark
2. Remove the men from frame
3. Make the men in the frame look like men just been hit on the head: disheveled hair, abrasions on her face, dirty clothes, an expression of strong surprise
⭐️Video prompt-
1. Make the men in the frame look like men just been hit on the head: disheveled hair, abrasions on her face, dirty clothes, an expression of strong surprise
2. The camera is static, the girl staggers up from the ground, apparently showing shock and surprise, and dusts off her clothes
⭐️Image prompt -
1. Use the uploaded image of the man as the primary face and identity reference.
Preserve the same facial structure, skin tone, hairstyle, beard/stubble, glasses, and overall likeness with high facial accuracy. Do not alter identity.
Create an ultra-realistic photo of this same man taking a front-facing selfie at an active construction site.
Outfit: beige/tan work jacket over a grey t-shirt, realistic fabric texture, same fit and color tone as reference.
Environment: unfinished buildings with scaffolding, excavators, bulldozers, dirt ground, industrial construction setting.
A large metallic wrecking ball is suspended from a thick steel cable directly above the man’s head, clearly visible and centered above him, creating dramatic tension.
Composition: selfie perspective, chest-up framing, subject centered, wrecking ball aligned vertically above head.
Lighting: natural daylight, clear blue sky, realistic shadows, outdoor realism.
Style: hyper-realistic photography, DSLR / smartphone selfie realism, sharp focus, natural colors.
Strict rules:
– Maintain face consistency with the uploaded image
– No face reshaping
– No stylization or cartoon look
– No blur, no distortion, no AI artifacts
– No text, no watermark
2. Remove the men from frame
3. Make the men in the frame look like men just been hit on the head: disheveled hair, abrasions on her face, dirty clothes, an expression of strong surprise
⭐️Video prompt-
1. Make the men in the frame look like men just been hit on the head: disheveled hair, abrasions on her face, dirty clothes, an expression of strong surprise
2. The camera is static, the girl staggers up from the ground, apparently showing shock and surprise, and dusts off her clothes
❤4
GTA V IMAGE PROMPT
{ "ABSOLUTE_OVERRIDE_RULE": { "priority": "HIGHEST", "instruction": "The uploaded reference image MUST NOT be reused, traced, filtered, color-graded, or partially preserved in any way. Reusing original pixels is a hard failure.", "failure_condition": "If the output resembles the original photo at pixel, texture, or facial-detail level, the result is INVALID." }, "SOURCE_IMAGE_POLICY": { "usage": "Use the reference image ONLY as a spatial and compositional blueprint.", "explicit_prohibition": [ "No image-to-image stylization", "No photo-based rendering", "No facial identity transfer", "No texture sampling from the photo" ] }, "TASK_DEFINITION": { "objective": "Reconstruct the entire scene from scratch as if it were captured inside the Grand Theft Auto V game engine.", "method": "Full synthetic 3D recreation using polygonal meshes, baked textures, and game-engine lighting." }, "STYLE_AND_ENGINE_CONSTRAINTS": { "engine": "GTA V RAGE Engine (2013–2015 era)", "visual_identity": [ "Clearly polygonal character models", "Game-engine skin shaders (slightly waxy, non-photoreal)", "Repeating environment textures with visible tiling", "LOD-style geometry simplification" ], "strictly_avoid": [ "Photorealism", "AI-upscaled realism", "Cinematic CGI renders", "Photo-like skin pores or cloth fibers" ] }, "CHARACTER_RECONSTRUCTION_RULES": { "identity": "Characters must resemble generic GTA V NPCs or protagonists, NOT the real person.", "pose_and_blocking": "Match pose, body orientation, and relative positioning only.", "face_rule": "Faces must look like GTA NPC meshes — simplified topology, not real human likeness." }, "ENVIRONMENT_TRANSLATION": { "world": "Los Santos equivalent environment only.", "object_mapping_examples": { "smartphone": "in-game iFruit phone prop", "street": "GTA asphalt texture with decals", "buildings": "Los Santos modular architecture assets", "vegetation": "GTA V foliage system" }, "background_rule": "The background must look like an open-world game map, not a photographed city." }, "LIGHTING_AND_RENDERING": { "lighting_model": "Real-time GTA V lighting and shadow system", "atmosphere": "Subtle in-game haze and distance fog", "camera": "Standard third-person or free-roam gameplay camera" }, "OUTPUT_SPECIFICATION": { "format": "4K in-game screenshot look", "ui": "NO HUD, NO minimap, NO UI", "focus": "Gameplay-level sharpness, visible textures, no cinematic blur" }, "NEGATIVE_PROMPT": [ "Original photo preserved", "Filtered photograph", "Same face as reference", "Hyper-real skin", "Cinematic movie still", "AI photo realism", "Loading screen illustration", "Concept art painting style" ] } , KEEP THE FACE ALL SAME , DONT CHANGE THE FACE
GTA VICE CITY IMAGE PROMPT
Create a STRICT 9:16 SCENE-LOCKED STYLE TRANSFER.
The uploaded image is a FIXED SPATIAL BLUEPRINT.
You must preserve object count, object type, exact position, scale, orientation, camera angle, horizon line, and depth perspective.
Do NOT add, remove, duplicate, relocate, resize, or invent any objects.
Only re-skin and re-mesh existing objects in place.
Scene Analysis (Mandatory)
First, segment the reference image into discrete object slots.
Detect and lock every visible object (buildings, people, vehicles, trees, poles, signs, ground, sky).
Each object becomes a LOCKED SLOT with fixed position, size, and orientation.
No object may exist outside these slots.
Identity Lock (Critical)
The main subject’s face, hair, body proportions, pose, and exact clothing silhouette must remain 100% identical to the original.
Convert the subject into a 2002-era low-poly digital character mesh.
No beautification, reinterpretation, or pose change.
{ "ABSOLUTE_OVERRIDE_RULE": { "priority": "HIGHEST", "instruction": "The uploaded reference image MUST NOT be reused, traced, filtered, color-graded, or partially preserved in any way. Reusing original pixels is a hard failure.", "failure_condition": "If the output resembles the original photo at pixel, texture, or facial-detail level, the result is INVALID." }, "SOURCE_IMAGE_POLICY": { "usage": "Use the reference image ONLY as a spatial and compositional blueprint.", "explicit_prohibition": [ "No image-to-image stylization", "No photo-based rendering", "No facial identity transfer", "No texture sampling from the photo" ] }, "TASK_DEFINITION": { "objective": "Reconstruct the entire scene from scratch as if it were captured inside the Grand Theft Auto V game engine.", "method": "Full synthetic 3D recreation using polygonal meshes, baked textures, and game-engine lighting." }, "STYLE_AND_ENGINE_CONSTRAINTS": { "engine": "GTA V RAGE Engine (2013–2015 era)", "visual_identity": [ "Clearly polygonal character models", "Game-engine skin shaders (slightly waxy, non-photoreal)", "Repeating environment textures with visible tiling", "LOD-style geometry simplification" ], "strictly_avoid": [ "Photorealism", "AI-upscaled realism", "Cinematic CGI renders", "Photo-like skin pores or cloth fibers" ] }, "CHARACTER_RECONSTRUCTION_RULES": { "identity": "Characters must resemble generic GTA V NPCs or protagonists, NOT the real person.", "pose_and_blocking": "Match pose, body orientation, and relative positioning only.", "face_rule": "Faces must look like GTA NPC meshes — simplified topology, not real human likeness." }, "ENVIRONMENT_TRANSLATION": { "world": "Los Santos equivalent environment only.", "object_mapping_examples": { "smartphone": "in-game iFruit phone prop", "street": "GTA asphalt texture with decals", "buildings": "Los Santos modular architecture assets", "vegetation": "GTA V foliage system" }, "background_rule": "The background must look like an open-world game map, not a photographed city." }, "LIGHTING_AND_RENDERING": { "lighting_model": "Real-time GTA V lighting and shadow system", "atmosphere": "Subtle in-game haze and distance fog", "camera": "Standard third-person or free-roam gameplay camera" }, "OUTPUT_SPECIFICATION": { "format": "4K in-game screenshot look", "ui": "NO HUD, NO minimap, NO UI", "focus": "Gameplay-level sharpness, visible textures, no cinematic blur" }, "NEGATIVE_PROMPT": [ "Original photo preserved", "Filtered photograph", "Same face as reference", "Hyper-real skin", "Cinematic movie still", "AI photo realism", "Loading screen illustration", "Concept art painting style" ] } , KEEP THE FACE ALL SAME , DONT CHANGE THE FACE
GTA VICE CITY IMAGE PROMPT
Create a STRICT 9:16 SCENE-LOCKED STYLE TRANSFER.
The uploaded image is a FIXED SPATIAL BLUEPRINT.
You must preserve object count, object type, exact position, scale, orientation, camera angle, horizon line, and depth perspective.
Do NOT add, remove, duplicate, relocate, resize, or invent any objects.
Only re-skin and re-mesh existing objects in place.
Scene Analysis (Mandatory)
First, segment the reference image into discrete object slots.
Detect and lock every visible object (buildings, people, vehicles, trees, poles, signs, ground, sky).
Each object becomes a LOCKED SLOT with fixed position, size, and orientation.
No object may exist outside these slots.
Identity Lock (Critical)
The main subject’s face, hair, body proportions, pose, and exact clothing silhouette must remain 100% identical to the original.
Convert the subject into a 2002-era low-poly digital character mesh.
No beautification, reinterpretation, or pose change.
❤3
Aspect ratio, camera angle, horizon line, and perspective are ABSOLUTELY LOCKED.
GTA Vice City (2002) Reskin Logic
For each locked object slot, replace only materials, textures, and mesh style with a GTA Vice City (2002) equivalent of the SAME object type.
Object Mapping Rules
Buildings: Replace with Vice City–style Art Deco buildings using the same footprint and height. No skyline changes.
People: Replace each person with a Vice City–era NPC in the exact same pose, position, and scale.
Vehicles: Replace each detected vehicle with a Vice City vehicle of the same category and orientation.
Vegetation: Replace each tree with a low-poly Vice City palm tree at the exact same trunk position and height.
Ground: Re-skin only the detected ground area using authentic Vice City asphalt, pavement, or sand textures.
Lighting & Color Lock
Apply Vice City–style lighting while preserving the original light direction, shadow placement, and time-of-day cues.
Sunset: Orange–purple 80s haze only (no shadow changes).
Night: Neon pink/cyan rim lighting only where existing light sources already affect the scene.
Technical Rendering Specs
Engine: RenderWare (2002 authentic look)
Style: Mid-to-low poly geometry, visible edges, flat shading
Textures: Low-resolution, early-2000s console quality
Hard Limits
No ray tracing
No modern global illumination
No depth of field
No motion blur
No cinematic post-processing
Strict Rejection Rules
No scene rebuild or Vice City map recreation
No object drift (no movement, rotation, resizing, or depth change)
No background or sky replacement
No UI elements (no minimap, HUD, subtitles, health bars, or text)
Negative Prompt
Random Vice City streets, generic Miami, invented buildings, extra NPCs, added cars, ultra-HD, realistic skin, cinematic lighting, bokeh, motion blur, new background, scene extrapolation
GTA Vice City (2002) Reskin Logic
For each locked object slot, replace only materials, textures, and mesh style with a GTA Vice City (2002) equivalent of the SAME object type.
Object Mapping Rules
Buildings: Replace with Vice City–style Art Deco buildings using the same footprint and height. No skyline changes.
People: Replace each person with a Vice City–era NPC in the exact same pose, position, and scale.
Vehicles: Replace each detected vehicle with a Vice City vehicle of the same category and orientation.
Vegetation: Replace each tree with a low-poly Vice City palm tree at the exact same trunk position and height.
Ground: Re-skin only the detected ground area using authentic Vice City asphalt, pavement, or sand textures.
Lighting & Color Lock
Apply Vice City–style lighting while preserving the original light direction, shadow placement, and time-of-day cues.
Sunset: Orange–purple 80s haze only (no shadow changes).
Night: Neon pink/cyan rim lighting only where existing light sources already affect the scene.
Technical Rendering Specs
Engine: RenderWare (2002 authentic look)
Style: Mid-to-low poly geometry, visible edges, flat shading
Textures: Low-resolution, early-2000s console quality
Hard Limits
No ray tracing
No modern global illumination
No depth of field
No motion blur
No cinematic post-processing
Strict Rejection Rules
No scene rebuild or Vice City map recreation
No object drift (no movement, rotation, resizing, or depth change)
No background or sky replacement
No UI elements (no minimap, HUD, subtitles, health bars, or text)
Negative Prompt
Random Vice City streets, generic Miami, invented buildings, extra NPCs, added cars, ultra-HD, realistic skin, cinematic lighting, bokeh, motion blur, new background, scene extrapolation
❤3👍1
PROMPT
Apply this prompt on this image.
Enhance the existing photo only — keep the original framing, angle, composition, and facial
proportions exactly the same.Do not crop, zoom, or convert to a close-up.Improve overall
image clarity and sharpness naturally, with special attention to eye detail while keeping the
full face and original framing intact.
Refine focus slightly for better definition without making it overly sharp.
Apply soft, natural cinematic lighting correction, preserving realistic shadows and depth
already present in the image.
Enhance natural skin texture with visible pores and fine details — no beauty smoothing, no
skin softening.
Remove digital noise and compression artifacts while retaining organic texture.
Improve dynamic range and contrast subtly, with muted highlights and rich mid-tones.
Maintain accurate, organic skin tones with realistic color balance.
Add very subtle cinematic film grain for a professional movie-still look.
Background should remain exactly as it is, only slightly cleaner if noisy — no artificial blur
added.
Preserve all facial features, expressions, hair strands, and natural imperfections.
Photorealistic result, professional camera look, high-quality enhancement only — no
stylization.
🚫 NEGATIVE PROMPT (CRITICAL)
close-up, zoomed-in, cropped face, changed framing, face alteration, reshaped features,
beauty filter, oversmoothed skin, plastic skin, waxy texture, AI face, cartoon, anime, CGI,
illustration, painting, artificial blur, fake bokeh, harsh lighting, flat lighting, over-sharpening,
halos, HDR look, distorted eyes, asymmetry, extra details, digital artifacts, smartphone filter
look.
RECOMMENDED SETTINGS (IMPORTANT) Image-to-Image Strength / Denoise: 0.15 – 0.30
CFG / Guidance: 5 – 7
Steps: 25 – 35
Face Restoration: OFF
Upscale: After final result only (2× max)
Seed: Fixed (for consistency)
Apply this prompt on this image.
Enhance the existing photo only — keep the original framing, angle, composition, and facial
proportions exactly the same.Do not crop, zoom, or convert to a close-up.Improve overall
image clarity and sharpness naturally, with special attention to eye detail while keeping the
full face and original framing intact.
Refine focus slightly for better definition without making it overly sharp.
Apply soft, natural cinematic lighting correction, preserving realistic shadows and depth
already present in the image.
Enhance natural skin texture with visible pores and fine details — no beauty smoothing, no
skin softening.
Remove digital noise and compression artifacts while retaining organic texture.
Improve dynamic range and contrast subtly, with muted highlights and rich mid-tones.
Maintain accurate, organic skin tones with realistic color balance.
Add very subtle cinematic film grain for a professional movie-still look.
Background should remain exactly as it is, only slightly cleaner if noisy — no artificial blur
added.
Preserve all facial features, expressions, hair strands, and natural imperfections.
Photorealistic result, professional camera look, high-quality enhancement only — no
stylization.
🚫 NEGATIVE PROMPT (CRITICAL)
close-up, zoomed-in, cropped face, changed framing, face alteration, reshaped features,
beauty filter, oversmoothed skin, plastic skin, waxy texture, AI face, cartoon, anime, CGI,
illustration, painting, artificial blur, fake bokeh, harsh lighting, flat lighting, over-sharpening,
halos, HDR look, distorted eyes, asymmetry, extra details, digital artifacts, smartphone filter
look.
RECOMMENDED SETTINGS (IMPORTANT) Image-to-Image Strength / Denoise: 0.15 – 0.30
CFG / Guidance: 5 – 7
Steps: 25 – 35
Face Restoration: OFF
Upscale: After final result only (2× max)
Seed: Fixed (for consistency)