GPT Image Prompt
Casual photorealistic iPhone-style vertical selfie using the active LoRA identity reference as the exact face-and-body authority. Preserve the same real young East Asian woman’s recognizable facial geometry, eye spacing, eye shape, nose, lips, jaw, cheek structure, skull proportions, dark tied-back hair with loose messy bangs, light natural skin tone, shoulder-to-head relationship, shoulder width, natural seated body proportions, torso-to-leg balance where visible, waist thickness, hip width where implied by seated posture, body mass distribution, and grounded adult physical presence. Do not slim, beautify, elongate, narrow, reshape, or substitute her body with a generic attractive version; her seated silhouette must remain reference-faithful.
She is seated in a dim office late at night, leaning back slightly in an office chair with her torso angled diagonally toward the phone. One arm extends out of frame holding the camera at arm length, while the other arm rests low and relaxed near her side. Her expression is subtle and alive: a soft camera-aware glance, relaxed eyelids, slightly parted lips, and a calm tired office-pause feeling, as if caught during a quiet late-night work break. Preserve the source-prompt human moment without copying any blank, stiff, dead-eyed, or mannequin-like expression from the reference.
She wears the same fitted white button-up office shirt with an open collar, visible buttons, natural wrinkles, realistic fabric tension over her reference body shape, low-light fabric softness, and believable compression from the seated pose. Dark lower clothing is only partially visible within the close vertical crop. The shirt must not reshape her waist, torso, shoulders, or seated silhouette away from the LoRA reference proportions.
The background is an ordinary cluttered workplace: desks, scattered papers, cabinets, office chairs, computer monitors, dark windows, small office objects, uneven spacing, and practical late-night workplace messiness. The environment must feel physically scaled around her, casually used, imperfect, and not staged, luxurious, cinematic, or overly clean. Maintain believable office geometry, natural object scale, realistic chair contact, grounded seated weight, attached shadows, and physical coexistence between her body and the office chair.
All expression, gesture, outfit, pose, camera angle, lighting, environment, smartphone realism, and visual imperfections remain subordinate to the active LoRA identity reference. The final subject must still be unmistakably the exact same LoRA woman in both face and body, not a similar-looking substitute, not a beautified replacement, not a generic East Asian office worker, and not a more polished or model-like version of the prompt.
Handheld iPhone-style vertical selfie from slightly above at close arm-length distance, mild wide-angle smartphone perspective, close crop, imperfect framing, slight tilt imbalance, casual asymmetry, realistic edge softness, no DSLR compression, no cinematic portrait depth, no professional lighting. Dim fluorescent office lighting with shadow compression, uneven smartphone auto-exposure, subtle highlight clipping on skin or shirt, low-light denoise, mild texture smearing, localized sharpness variation, realistic skin shine, imperfect hair reconstruction, subtle compression artifacts, ordinary mobile HDR behavior, and natural smartphone processing flaws.
Avoid cinematic framing, influencer beauty, fashion-editorial posing, luxury-office staging, perfect symmetry, over-clean HDR, poreless plastic skin, over-sharpened detail, exaggerated bokeh, seductive gaze, posed smile, frozen neutral mouth, blank stare, dead eyes, mannequin posture, artificial attractiveness optimization, waist slimming, torso narrowing, leg elongation, hip narrowing, thigh reduction, body mass redistribution, generic model body, fake office geometry, floating contact, disconnected shadows, impossible chair interaction, fused fingers, duplicated fingers, melted hands, broken wrist angles, and any smartphone perspective that changes the LoRA reference body identity.
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”.