A four-beat skincare vlog prompt built for identity lock

This prompt scripts a 15-second vertical selfie clip of a young woman applying a named skincare product, broken into four timed beats with fixed dialogue and lighting. It suits anyone testing whether a model can hold a face, a product label and a lighting condition steady across a short handheld sequence.

Model: GPT Image 2

iPhone selfie photo, mid angle slightly below eye level, of @image1: the same young woman as the reference photo, oval face, tan-fair skin, brown eyes, full lips, long straight-to-wavy light brown hair with soft caramel-blonde balayage, center part. She wears a blue and white floral midi skirt with a fitted white ribbed top, and a loose open white linen shirt worn over it as a light layer, Tom Ford black aviator sunglasses pushed up on her head, blue resin C-hoop earrings, and a baroque pearl pendant necklace, sitting at a sun-drenched Mediterranean outdoor cafe terrace, terracotta clay pots with bougainvillea behind her, a whitewashed stone wall edge in frame, dappled olive tree shadow playing across her face and shoulders, warm golden afternoon light at color temperature #E8A85C, a neutral cool editorial expression, slight medium background blur, an effortless relaxed confident attitude. Face and hair 100% match the reference. Fully clothed, modest coverage throughout, no swimwear, no exposed midriff.

A 15-second vertical (9:16) ultra-photorealistic close-up vlog clip, filmed on a phone held at arm's length. Not an ad. A girl talking to her own camera at the end of the day, the way people actually film themselves.
TALENT —
Young woman, early-to-mid 20s. Long light-brown hair with soft caramel balayage, middle part, loose undone waves, a few strands stuck to her cheek. Fair-to-light-tan skin with visible freckles across the nose and cheeks, thick natural brows, blue-green eyes, bare face, small blue resin hoop earrings, thin gold chain with a small pearl pendant. Face, freckles, hair and eye color must 100% match the reference. Oversized white linen button-down over a fitted white ribbed top, fully covered, collar slightly crooked, sleeves pushed up.
PRODUCT —
LANEIGE Water Sleeping Mask_EX, 70 ml. Frosted translucent light-blue glass jar, smooth matte cobalt-blue screw lid. Front label reads, in this exact order: "LANEIGE" in wide-spaced capitals, "WATER SLEEPING MASK_EX" below it, then "SLEEPING MICROBIOME" and "BRIGHTENING · HYDRATING · SOFTENING" in smaller pale-yellow type, then "70 ml" at the base. Inside, thick pale mint-blue gel with a wet glossy surface that holds a soft peak. Shape, color, gel texture and all label text must 100% match the reference, correctly spelled, legible whenever in focus.
LOCATION
A villa bedroom at golden hour. Only fragments read in frame because the camera is so tight: white stucco wall, a corner of rumpled white linen bedding, blurred bougainvillea and olive leaves through an open door behind her. The background stays soft and mostly out of focus the entire time.
LIGHTING
One condition, unchanged across all four beats. Low warm sun from frame-left through the open door, roughly 4800K, soft white-wall bounce filling the shadow side of her face. Warm rim on the left edge of her hair and jaw. Slight blown highlight where the sun clips her cheek. No ring light, no fill, no glam lighting.
FRAMING
Close-up throughout. Her face fills most of the frame in every beat, chin to just above the hairline, never wider than upper-chest. The jar enters frame close to her face, never held out toward the lens.
0:00–0:04
Visual: Tight front-camera close-up, slightly below eye level. She's mid-motion already, pushing hair off her face with the back of her wrist. She glances just off-lens, checking her own image, then back. The jar comes up beside her cheek, held low and casual, only the top half in frame, lid already off. A slow exhale through the nose.
Camera: Selfie camera, arm's length, natural handheld drift, one quick autofocus hunt between her face and the jar.
Dialogue: "Okay. Last thing, then bed."
0:04–0:07
Visual: She tips the jar slightly and drags two fingertips through the gel. Macro-tight: her fingertips, the wet mint-blue surface, a fine thread of gel stretching and snapping. The "LANEIGE" and "WATER SLEEPING MASK_EX" lines cross the frame in focus for a beat. Her face stays partly in frame at the edge, out of focus.
Camera: Same front camera pushed in close, shallow focus, rack from the label to the gel on her fingers.
Dialogue: "It's cold, by the way."
0:07–0:11
Visual: Back to her face, closer now. She taps the gel across one cheek, her forehead, her chin, then presses it in flat-palmed, working outward. The gel flashes milky over her skin, then settles clear with a faint sheen. She winces at the cold, blinks it off, keeps going. Freckles, pores and natural under-eye shadows stay visible right through the layer. She's talking while working, not pausing for the line.
Camera: Selfie camera, very close, small handheld corrections, a slow drift as she turns her cheek toward the light.
Dialogue: "Thin layer. You don't wash it off."
0:11–0:15
Visual: Closest framing of the clip. She presses two fingertips against her cheek and lets them lift, feeling the faint tack. Turns her head slowly through the low sun, checking the sheen along her cheekbone, eyes tracking her own reflection in the screen rather than the lens. A short nod. The corner of the blue jar sits soft-focus in the lower frame edge, label partly readable. She reaches to stop the recording and the frame tilts as her hand moves.
Camera: Selfie camera, slow handheld orbit from three-quarter profile to straight-on, ending on her face filling the frame, then a small tilt as she reaches.
Dialogue: "Morning me is gonna be so happy."
PERFORMANCE
End-of-day, low energy, unperformed. She talks the way someone talks while doing something else, lines half-thrown-away, not projected. Micro-behavior carries the beats: a slow blink, a genuine flinch at the cold, a half-smile that arrives late, eyes checking her own reflection instead of the lens, one natural swallow, visible breath, a small unconscious head tilt. No presenter voice, no bright delivery, no exaggerated reactions, no pointing at the product.
VOICE
Soft, slightly flat, close to the mic because the phone is at arm's length. Natural room tone around it. Unhurried, a little tired. Ordinary speaking volume with a touch of breath on the ends of words.
REALISM
Pore-level skin realism: visible pores across the nose and cheeks, fine forehead texture, natural freckling, faint under-eye shadows, a small blemish near the jaw, stray flyaway hairs catching the backlight, vellus hair on the cheek edge. Wet living eyes with catch-lights and natural asymmetric blinking. The gel reads as a real cosmetic layer refracting light on top of her skin, never replacing its texture.
PHYSICS
Gel holds a soft peak on her fingertip and slumps under pressure. Hair moves with real inertia and a few strands fall back across her face. Linen creases and drapes naturally. The jar has real weight and correct contact shadow against her hand. Camera weight shifts naturally with her arm.
CAMERA CHARACTER
Authentic modern smartphone selfie footage. Natural handheld tremor throughout, small framing corrections, quick autofocus hunts, mild wide-angle distortion at close range, natural rolling shutter, real motion blur, slight sensor noise in the shadow side. Never feels stabilized or operated by anyone else.
STYLE
Real personal vlog, iPhone-shot look, high-end photorealistic quality. Warm golden-hour daylight, real ambient evening atmosphere, no retouching, no skin smoothing, no product-hero framing, no before/after transformation, no commercial polish.
OUTPUT SETTINGS
9:16 vertical. 15 seconds. Real-time motion. High-detail photorealistic 8K intent. Natural smartphone motion blur and subtle sensor noise.
POSITIVE LOCKS
 keeps the same face, freckles, hair, earrings, necklace and white linen shirt over white ribbed top in every beat, fully covered throughout.
 stays the same frosted blue jar with matte cobalt lid, correct one-hand size, label text correctly spelled and legible whenever in focus, sharpest in the 0:04–0:07 beat.
Exactly two arms and two hands visible, anatomically correct with five fingers, one hand always holding the phone.
Same room, same wardrobe, same golden-hour lighting from first frame to last.
Close-up framing held throughout, her face dominant in every beat.
Spoken dialogue exactly as written, in order, at the stated timings, with accurate natural lip-sync.
Natural ambient audio only: faint breeze, distant birds, the small click of the lid, room tone. No music, no on-screen text, no captions, no subtitles, no logos other than the product's own packaging.