Create

Generate Similar

Generation Mode Choose how the image should start.

Prompt

Generate an ultra-photorealistic vertical 9:16 selfie. Use the TWO uploaded face references (the woman and the man) and preserve their identities strictly 1:1: 100% likeness. Do not alter facial features, proportions, age, eye shape, nose, lips, eyebrows, or skin tone. The faces must be fully recognizable, with no beautification or enhancement.

Scene & Pose

The woman and the man are sitting side by side on the edge of a metal grated platform at an extreme height on the roof of a skyscraper. Both are looking directly into the camera with a subtle, confident smile.

Required pose details:

Relaxed posture
Natural weight shift
Asymmetrical pose (shoulders at slightly different heights, natural spinal curve, torso slightly rotated)
Hands and fingers must be anatomically correct

Camera & Perspective

iPhone selfie aesthetic
Wide-angle lens (~24mm equivalent)
Slight front-camera perspective distortion
Camera positioned slightly above eye level
Faces should appear large in the frame (head and shoulders), while clearly showing:
the edge of the platform,
the steel grating beneath them,
and the depth visible through the grid with the city far below, making the height immediately obvious.
Do not blur the background—the cityscape should remain clearly readable.

Location (describe as accurately as possible)

The edge of a skyscraper rooftop with a narrow galvanized open steel grating service platform featuring square grid openings.
Visible steel ribs, edge profile, bolts, nuts, washers, and mounting plates.
Right beside the couple is a technical equipment box with:
a control panel,
several round buttons (red, black, metallic),
toggle switches,
terminals,
corrugated cable conduits.
No readable text or labels.
Nearby are two red aviation warning beacon lights mounted in metal bases.
Behind them stands a tall lattice antenna tower with steel diagonal braces, mounting joints, and cables extending upward beyond the frame.
Background:
sprawling metropolitan skyline,
tiny buildings far below,
light atmospheric haze,
distant horizon,
clear sky,
daylight or soft golden-hour lighting,
a powerful sense of extreme height.

Clothing & Accessories (UPDATED — match the references exactly)

Woman

Minimalist strapless black top, exposing shoulders and collarbones.
Bottom: entirely black in a total black aesthetic, wearing high-waisted black pants matching the reference (dense fabric or leather).
Headwear:
black ribbed knit beanie with large cat ears, exactly like Reference 2.
The beanie must sit naturally on the head.
Do NOT pull it over the face.
The face must remain completely visible.
Jewelry:
large silver hoop earrings,
multiple ear hoops/cuffs matching Reference 2,
tight black choker around the neck.
No face masks or face coverings.

Man

Black ribbed tank top (“wife beater” style), fitted exactly like Reference 1.
Visible ribbed fabric texture and natural wrinkles.
Black pants in a total black aesthetic.
Accessories:
no cap,
no bandana,
no face covering,
face and hair completely visible.
If gloves or hand accessories appeared in the reference, do not include them unless absolutely essential.

Quality & Photorealism

Photo RAW style
Sony A7R look
Ultra photorealistic
Extremely high detail
Sharp focus on the faces
Natural color rendering
Realistic shadows
Accurate contact shadows beneath the bodies and legs on the steel grating
No CGI appearance

Negative Prompt

face change, identity drift, different makeup, different hair, any face covering on either person, balaclava, mask, bandana, cap or hat on the man, hat covering the woman’s face or eyes, outfit change, wrong accessories, extra fingers, extra limbs, warped hands, floating, mismatched shadows, blur, plastic skin, CGI look, text, watermark, UI elements, logos.

3764 / 10000

Based on: Couple sits on skyscraper edge with city far below

Model

Select a model to continue.

Each model exposes different quality and editing capabilities.

Size & Aspect Ratio

Pick the output framing.

Available options change with the selected model.

Select a model to see credit cost. Reference mode adds 1 credit on supported models.

Generate Similar

Ready when you are

Generating your image

Image ready

We could not finish this image

Generating your image

Image ready

We could not finish this image

Send people to the full prompt and image