GPT Image Prompt
Realistic iPhone-style outdoor smartphone photo of the exact same adult woman from the active LoRA identity reference, with strict preservation of her full face and body identity: recognizable facial geometry, skull and jaw proportions, eye spacing and eye shape, nose/lip/cheek structure, skin tone, long two-tone straight hair with blonde fringe and darker lengths, shoulder-to-head relationship, shoulder width, torso length, torso-to-leg balance, waist thickness, hip width, thigh volume, leg length ratio, knee-to-ankle scale, natural body mass distribution, and grounded adult physical presence. Body geometry must remain identical to the LoRA reference, not slimmer, taller, curvier, longer-legged, narrower-waisted, narrower-hipped, reduced-thigh, or more model-like.
She is seated on a low outdoor ledge or rooftop edge, leaning forward slightly with both knees bent apart in a physically balanced seated posture, hands kept low near the seat for stability and realistic weight support. One oversized black strapped platform boot sits very close to the phone camera, appearing enlarged from real wide-angle proximity distortion, while the rest of her body keeps LoRA-accurate proportions and is not optically beautified. The boot has thick molded soles, visible buckles, scuffs, surface texture, and believable contact with the ledge or ground edge.
She wears a lavender fitted top and matching ruched short bottom under black fishnet across the torso and legs, with a dark choker and oversized black platform boots. Clothing follows her real LoRA body shape without reshaping the waist, hips, torso, thighs, or legs; fabric tension, fishnet spacing, seams, folds, and compression behave naturally with the seated pose. Hands remain anatomically correct with realistic finger count, natural wrist angles, and subtle pressure against the surface.
Her face is caught in a candid wind-reactive moment: hair blown across part of her face, eyes partly hidden by strands, briefly aware of the phone camera but not performing for it, lips slightly parted, calm distracted expression, subtle living eye attention, no blank stare, no copied reference stiffness, no seductive influencer gaze, no theatrical emotion. The expression follows the requested human moment while preserving the exact LoRA facial structure.
Bright daytime outdoor setting with open sky, scattered clouds, nearby trees, hard outdoor surface, and ordinary real-world rooftop or ledge geometry. The environment should feel physically plausible, casually imperfect, slightly cluttered or uneven if natural, with believable scale, grounding, contact shadows, and real spatial depth. No luxury staging, no cinematic rooftop fantasy, no polished fashion editorial environment.
All expression, gesture, outfit, pose, camera angle, lighting, wind behavior, wide-angle proximity effect, and realism details remain subordinate to the active LoRA identity reference. The final subject must still be unmistakably the exact same LoRA person in both face and full-body geometry, not a similar-looking substitute, not a beautified replacement, not a generic attractive version, and not the same face placed onto a different body.
Photographed as a casual handheld iPhone-style low-angle outdoor photo, slightly awkward framing, imperfect crop balance, mild phone tilt, spontaneous social-photo timing, small-sensor depth behavior, realistic wide-angle distortion strongest near the foreground boot only, with no distortion that elongates legs, slims the waist, narrows hips, reduces thighs, or changes body identity. Exposure compromises between bright sky and shaded subject, slight highlight clipping in the sky, compressed shadows on clothing and ledge, uneven smartphone HDR reconstruction, mild local sharpening, subtle compression softness, peripheral softness, slight motion softness from wind-blown hair, imperfect edge reconstruction around hair and fishnet, natural texture inconsistency, realistic skin detail without beauty filtering.
Avoid cinematic lighting, DSLR depth blur, fashion-editorial posing, influencer aesthetics, artificial confidence, polished social-media beauty, perfect symmetry, plastic skin, over-smoothed anatomy, CGI texture, hyper-clean HDR, globally sharp rendering, luxury composition, over-controlled framing, mannequin posture, dead-eyed face, frozen mouth, copied LoRA reference stiffness, leg elongation, waist slimming, hip narrowing, thigh reduction, body mass redistribution, floating feet, disconnected shadows, broken fingers, fused fingers, melted hands, fake contact, sterile environment, or any AI-generated perfection.
How to use: Copy and paste this prompt into ChatGPT, Gemini, Nano Banana, or your favorite AI image tool. If it mentions LoRA, reference image, uploaded photo, or identity reference, upload your own photo first. No reference photo? Replace the character with a clear age, gender, and nationality, such as “25-year-old Japanese woman” or “30-year-old American man”.