Testing ChatGPT Text to Image: Two Prompts

AI (ChatGPT) is getting better, obviously. What is suprising is that each time it generates a different image even with the same prompt!

Full-body, photorealistic editorial of a young East Asian woman in a bright outdoor corridor with sunlit brick archways in the background; cheerful, breezy mood. She stands side-on in left profile close to a light concrete wall. Left arm extended forward with relaxed wrist, fingertips touching the wall at about shoulder height for a gentle lean; right arm bent with the hand resting lightly on the high-waist/hip seam (thumb tucked, elbow slightly back). Torso upright with a soft S-curve, head turned back toward the camera, calm smile and confident gaze. Legs long and elegant: feet in pointed heels, front leg straight and slightly forward, back leg half-step behind with the knee softly bent and heel subtly lifted, ankles close so the legs form a graceful diagonal toward camera. Wardrobe: airy white lace off-shoulder blouse, pale pastel high-waisted tailored shorts, white stilettos, minimal jewelry, neat bob hair, refined natural makeup. Light & setting: clean midday sun with soft bounce fill; gentle rim light from the open archway; shallow depth of field so the brick arches and passage blur pleasantly; no other people. Aesthetic: high-fashion Vogue editorial, cinematic elegance inspired by Annie Leibovitz and Zhang Jingna; realistic skin texture, accurate proportions, crisp fabric detail; ultra-real, 8k, magazine cover quality. Negative prompt: extra limbs, deformed hands, warped ankles, twisted waist, stiff pose, heavy makeup, messy hair, garish colors, text or watermark, crowds, low-res, oversharpening, blown highlights, harsh shadows, plastic skin.

Full-body, photorealistic editorial of the given girl in the photo, She is crouched beside a shallow round stone water basin with a small tabby cat sipping from the rim. Pose (match closely): camera 3/4 left. She is in a deep squat/kneel: right knee down, left foot flat and knee forward, hips close to heels. Torso folds slightly from the hips toward the basin, shoulders relaxed. Left forearm rests lightly across the left knee, right hand touches the stone rim near the cat. Head tilted down toward the water, eyes following the cat with a soft smile; chin slightly tucked, fringe framing a short straight bob. Show clear reflections of both her face and the cat on the water surface. Styling: light summer outfit (sleeveless ivory top, high-waist pale linen shorts), minimal jewelry, natural luminous makeup, healthy skin glow. Environment & mood: rustic wooden house and vines in the background, dappled late-afternoon sunlight, warm highlights, cool shadows; fresh green foliage; gentle breeze ripples. Overall joyful, tender, quiet-summer atmosphere. Lighting & aesthetic: soft natural key with subtle bounce fill, slight back rim lighting on hair; shallow depth of field (subject/cat/water sharp, background softly blurred). High-fashion Vogue mood with editorial color grading in the spirit of Annie Leibovitz and Zhang Jingna; ultra-detailed textures of skin, stone, fur, and water; 8k, clean composition. Negative prompt: childlike features, extra limbs, bad hands, distorted crouch, stiff pose, wet clothes, harsh shadows, blown highlights, clutter, multiple people, text, watermark, low-res, plastic skin, oversharpening, cartoon/anime style.