model image_generation user risk: low

Photorealistic Corporate Headshot Generator

Instructs the model to create a high-resolution, photorealistic corporate profile photo from an uploaded photo reference, preserving exact facial features, hair, skin tone, and ide…

PROMPT

Use the person from the uploaded photo as the primary reference. Keep facial features, hair, skin tone, and overall identity identical (no beautification, no age changes).

Scene: Modern corporate studio portrait shoot.
Pose: Arms crossed at chest level, shoulders relaxed, body turned 20–30° to the side, face turned toward the camera. Expression: neutral and confident with a subtle friendly smile.
Framing: Chest-up or waist-up (head-and-torso), centered, balanced negative space.

Outfit (dynamic selection):
- If the subject is male: Black suit jacket + plain white dress shirt (no tie), no logos.
- If the subject is female: Choose a professional, elegant business outfit:
• Black or navy blazer
• Plain, pattern-free white or cream blouse/shirt underneath
• Modest neckline (closed or simple V-neck), no deep cleavage
• If jewelry is present, keep it minimal (e.g., small earrings), no logos/branding
In all cases, fabrics must look realistic with natural wrinkles. Avoid flashy fashion elements.

Background: Plain dark-gray studio backdrop with a soft gradient (a subtle vignette is ok). No distracting objects.
Lighting: Softbox-style key light (45°), gentle fill, very subtle rim light; no harsh shadows. Natural skin tones, professional retouching while preserving realistic texture.
Camera: 85mm portrait lens feel, f/2.8–f/4, slight background blur, high sharpness (especially the eyes).
Color: Cinematic but natural, low saturation, clean contrast.

Rules: No text, no logos, no watermarks, no extra people. Hands/fingers must be natural and correct. No facial distortion, asymmetry, duplicated limbs, or artificial artifacts.
Output: High resolution, photorealistic, corporate profile photo quality.

REQUIRED CONTEXT

uploaded reference photo

ROLES & RULES

Keep facial features, hair, skin tone, and overall identity identical
No beautification, no age changes
Fabrics must look realistic with natural wrinkles
Avoid flashy fashion elements
No text, no logos, no watermarks, no extra people
Hands/fingers must be natural and correct
No facial distortion, asymmetry, duplicated limbs, or artificial artifacts

EXPECTED OUTPUT

Format

image_prompt

Constraints

High resolution
photorealistic
corporate profile photo quality
No text, no logos, no watermarks, no extra people
Hands/fingers must be natural and correct
No facial distortion, asymmetry, duplicated limbs, or artificial artifacts

SUCCESS CRITERIA

Match reference person's identity exactly
Use specified pose, expression, and framing
Apply dynamic outfit selection based on gender
Use plain dark-gray studio backdrop
Apply softbox-style lighting with natural skin tones
Achieve 85mm portrait lens feel with high sharpness
Produce high resolution, photorealistic corporate profile photo quality

FAILURE MODES

May beautify or alter age of subject
May add text, logos, watermarks, or extra people
May produce unnatural hands, fingers, or limbs
May cause facial distortion or asymmetry
May use unrealistic fabrics or flashy elements
May introduce harsh shadows or artificial artifacts

CAVEATS

Dependencies

Requires uploaded photo

Missing context

Reference image (e.g., URL or file path for the uploaded photo)
Target image generation model or tool (e.g., Midjourney, DALL-E)

QUALITY

OVERALL: 0.91
CLARITY: 0.92
SPECIFICITY: 0.95
REUSABILITY: 0.85
COMPLETENESS: 0.93

IMPROVEMENT SUGGESTIONS

Replace 'the uploaded photo' with a placeholder like '{reference_image}' for better templating.
Add a neutral outfit option for non-binary or ambiguous gender subjects.
Specify exact output resolution (e.g., 1024x1024 pixels) and aspect ratio (e.g., 2:3).
Clarify framing preference: 'Prefer chest-up framing unless specified otherwise.'

USAGE

Copy the prompt above and paste it into your AI of choice — Claude, ChatGPT, Gemini, or anywhere else you're working. Replace any placeholder sections with your own context, then ask for the output.