GPT Image Prompt
Same young adult woman as the active LoRA identity reference, preserving her exact LoRA facial identity and body identity: same face geometry, eye spacing, nose-lip-jaw relationship, cheek structure, hair identity, skin tone, shoulder width, shoulder-to-head relationship, torso proportion, waist thickness, hip impression, and overall physical presence. LoRA identity is the authority for both face and body. Do not substitute a similar-looking woman, do not beautify her into a different person, do not slim the waist, narrow the hips, reduce thigh volume, lengthen the legs, or turn her into a model-like or influencer-style subject.
A nostalgic 1990s disposable-camera flash photograph taken inside a small, busy Japanese izakaya. The setting feels cramped, lively, and authentic, with handwritten menu cards, retro posters, worn wall details, and the visual clutter of a real neighborhood bar. The young woman sits on a wooden counter stool and turns her upper body back toward the camera in a natural, casual seated twist. She wears a soft red cardigan layered over a white tank top, paired with a short black skirt. The red cardigan should pop vividly under the direct flash and become a strong visual anchor in the image.
Her expression should feel emotionally alive, intimate, and natural rather than posed: relaxed but attentive eyes, a soft human gaze toward or slightly past the camera, lightly flushed cheeks, and smooth glossy lips with a subtle lived-in softness. Preserve a candid, imperfect human moment, not a mannequin face, not a fashion smile, and not a dead-eyed stare. Her posture should feel casually seated and physically believable, with natural shoulder asymmetry, realistic torso twist, and ordinary seated balance.
Behind her, older diners occupy the darker background, softly visible within the dim, noisy atmosphere of the izakaya. They should feel like real people in a lived-in environment, helping create contrast between the flash-lit subject and the murkier, ambient background. The scene should feel intimate, nostalgic, slightly chaotic, and emotionally warm, with believable environmental depth and real-world clutter rather than clean composition.
All expression, pose, outfit, lighting, environment, and camera behavior remain subordinate to the active LoRA identity reference. The final image must still be unmistakably the exact same LoRA subject in both face and body, not a similar-looking substitute, not a beautified replacement, and not a generic attractive version of the prompt.
Camera and realism behavior: simulate a real 1990s disposable-camera snapshot with harsh direct on-camera flash, gritty film grain, visible noise, strong shadow falloff, blown highlight hotspots, mild color inconsistency, imperfect white balance, and casual handheld framing. Preserve the raw, spontaneous look of a one-time flash capture: slightly awkward crop balance, mild framing imperfection, subtle motion instability, uneven sharpness, and authentic snapshot energy. Avoid cinematic grading, editorial polish, luxury fashion energy, perfect symmetry, artificial bokeh, clean studio lighting, or over-optimized beauty rendering. Skin should retain realistic texture and slight imperfection; the environment should feel physically real, cluttered, and socially believable. Overall result: a candid, emotionally intimate, imperfect flash snapshot with strong nostalgic 90s Japanese izakaya atmosphere.
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”.