Strawberry Milkshake Diptych AI Photo Prompt for Gemini
Generate a photorealistic editorial-style diptych where two uploaded portraits of the same person interact through a central strawberry milkshake. The composition uses a clear plastic cup positioned precisely on the frame edge to create a seamless illusion. The lower subject reaches up, holding the condensation-covered cup, while a straw extends directly to the pursed lips of the upper subject. Both faces remain exact, unaltered replicas of the reference images, featuring natural skin texture and bright sunlight. One subject wears an oversized yellow cat-print tee and rimless glasses with a messy bun, while the other sports a green 'CATS' graphic tee, parted hair, and glossy black sunglasses. The blurred background features a tall black brick building, enhancing the clean 9:16 vertical lifestyle aesthetic.
Generated Images
The Prompt
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, bright sunlight, and a lifestyle editorial aesthetic. The central focus of the composition is a single glass, which connects both frames and creates a visual illusion. A clear plastic cup with a strawberry milkshake sits on the edge between the upper and lower frames. The person below, in the first photo, extends her/his arm upward from the lower frame and holds the cup at waist level. the person wears a thin bracelet on her/his wrist. Her/his hand passes through the space between the frames, as if she/he is pouring the drink upward. A light-colored straw emerges from the cup, pointing toward the upper frame directly at the lips of the person above. The person above, in the second uploaded photo, drinks through the straw, her/his lips slightly pursed, creating the impression that two versions of the same person are interacting. The glass is filled with a strawberry milkshake, with condensation and light reflections visible on the transparent walls. This element is positioned right in the center of the composition, uniting both images into a single creative shot. Image: Top shot: the person from the first image , voluminous, and casual bun, with soft waves and loose strands framing her/his face. She/he wears narrow, transparent, rimless glasses. Clothing: An oversized yellow crew-neck T-shirt with short, elbow-length sleeves. On the chest, a minimalist cat print. Bottom shot: the person from the 2nd image parted in the middle. Large, glossy black sunglasses. Lips pursed in a kissing pose. She/he wears a bracelet/watch on her/his wrist. She/he wears an oversized green urban-style T-shirt with a large black and white print on the chest: the word "CATS" underneath. It is loose-fitting and has short sleeves. background: An elegant, tall, black brick building, slightly out of focus. RA. 9:16.
How to Use
To use this prompt effectively, you will need to upload two clear, high-resolution reference images of the same person to Gemini AI or a similar platform that supports face-consistency features. First, ensure your reference images have good, even lighting on the face to achieve that bright sunlight editorial look. Paste the prompt exactly as provided into the text field. If the AI struggles with placing the glass perfectly on the seam between the two images, you can add emphasis to the framing by specifying 'exact vertical split screen' or 'hard frame break.' You can tweak minor details like the flavor of the milkshake by changing 'strawberry' to 'chocolate' or 'vanilla,' but keep the color contrast in mind against the yellow and green shirts. For better results regarding the face replication, make sure the uploaded photos show the face clearly without heavy shadows or extreme angles. The prompt is optimized for a 9:16 vertical aspect ratio, so set your output dimensions accordingly to maintain the intended editorial composition and avoid cropping the building background or the interacting hands.
Prompt Variations
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, soft window light, and a corporate editorial aesthetic. The central focus is a single glass of iced matcha latte with a bamboo straw, sitting precisely on the edge between frames. The person below extends an arm upward from the lower frame, holding the sweaty glass at chest level. Their hand passes the frame break. A bamboo straw points directly at the lips of the person above, who is sipping gently. Upper subject: from the first image, sleek low ponytail, gold wire-frame glasses, and an oversized beige blazer over a plain white crewneck tee. Lower subject: from the second image, hair down with a slight wave, minimal makeup, wearing a black slim-fit turtleneck. Background: A softly blurred modern office with a standing desk and green plant. 9:16.
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, harsh midday sun, and a vibrant poolside aesthetic. The central focus is a clear plastic cup filled with bright blue slushie, placed exactly on the frame line. The lower subject reaches up, hand visible with a colorful beaded bracelet, holding the cup from below. The neon straw curves upward to the lips of the upper subject, who is drinking with a slight smile. Upper subject: from the first image, hair pulled back in a messy wet look, wearing a loud floral short-sleeve button-up shirt open over a swim top. Lower subject: from the second image, hair slicked back, wearing a simple white ribbed tank top. Background: A slightly out-of-focus swimming pool with turquoise water and sun reflections on the surface. 9:16.
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, warm golden hour backlight, and a nostalgic street style aesthetic. The central object is a plastic iced coffee cup with heavy condensation and a white straw, positioned on the frame border. The lower person's hand extends upward, holding the cup from the bottom. The straw points up to the lips of the upper person who is taking a sip. Upper subject: from the first image, voluminous curls, wearing a vintage wash denim jacket with the collar popped over a white tank. Lower subject: from the second image, center-parted hair, wearing an oversized rust-colored hoodie with the hood down. Background: A blurry city street at sunset with warm bokeh light orbs. 9:16.
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, dappled sunlight filtering through leaves, and a fresh garden editorial aesthetic. The central connecting element is a tall glass of sparkling lemonade with a lemon wheel on the rim, placed on the edge between frames. The lower subject's arm reaches up, holding the base of the glass. The clear straw extends directly to the upper subject's lips. Upper subject: from the first image, a loose low bun with wispy baby hairs, wearing a delicate floral print sundress. Lower subject: from the second image, hair tucked behind ears, wearing a relaxed oatmeal-colored linen button-up shirt. Background: A dense, out-of-focus green hedge and climbing ivy. 9:16.
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, soft neon ambiance mixed with window light, and a retro diner aesthetic. The central focus is a fountain glass with a cola float, vanilla ice cream scoops, and a bright red straw positioned on the frame seam. The lower person's hand, adorned with a silver ring, holds the glass from the bottom edge. The straw leads up to the drinking lips of the upper person. Upper subject: from the first image, hair in a high pony with a scrunchie, wearing a retro red striped polo shirt. Lower subject: from the second image, messy hair, wearing a faded vintage band t-shirt. Background: A blurry diner counter with chrome accents and a neon 'OPEN' sign glowing softly. 9:16.
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, cool overcast light from a window, and a cozy introspective aesthetic. The central element is a ceramic mug of hot tea placed exactly on the horizontal split. The lower subject holds the mug from below, hands visible. Instead of a straw, a delicate wisp of steam rises vertically, curling directly toward the nose and slightly parted lips of the upper subject who is blowing gently on the tea. Upper subject: from the first image, soft messy bun, wearing a chunky cable-knit cream sweater. Lower subject: from the second image, hair down, wearing a soft gray wool scarf draped loosely. Background: A slightly blurred windowpane covered in raindrops. 9:16.
create a image of 2 person from the uploaded image , without altering their faces, making them exact replicas of the reference image. A photorealistic diptych of two horizontal shots, with great skin detail, natural facial texture, cool neon night lighting, and a futuristic urban aesthetic. The central focus is a clear plastic cup of brown sugar bubble tea with tapioca pearls visible at the bottom, sitting on the frame edge. The lower subject's hand, wearing a matte black techwear glove, holds the cup from below. The thick straw points directly at the pursed lips of the upper subject. Upper subject: from the first image, sleek center-parted hair, wearing reflective silver shield sunglasses and a black high-neck technical jacket. Lower subject: from the second image, hair slicked back, wearing a matte black utility vest over a long-sleeve mesh top. Background: A blurry, rain-slicked alleyway with neon blue and purple lights. 9:16.
FAQ
- Why does the AI sometimes change the face slightly even when the prompt says not to?This is a common limitation with current AI image models regarding 'face consistency' across different poses and angles within a single generation. Even with a direct instruction to use the uploaded image as an exact replica, the model interprets the scene as a whole and may make micro-adjustments to fit the new lighting, angle, or expression described (like pursed lips). To improve accuracy, ensure the face is the largest, clearest part of your reference uploads. You may also need to use an additional 'inpainting' or 'face swap' tool after generation to restore the exact original features if the likeness drifts.
- The glass is floating between the frames instead of looking like it's held. How do I fix this?This illusion relies heavily on the AI understanding the exact framing boundary. If the glass appears to float in a gap, try regenerating and adjusting the prompt wording slightly. You can add clarifying terms like 'the bottom half of the cup rests exactly on the bottom edge of the top frame' or 'the seam between the two photos cuts directly through the middle of the milkshake cup.' Additionally, ensure your output aspect ratio is locked to 9:16. If the AI still creates a visible gutter, you can manually crop the images together in an external editor to achieve the perfect seamless line.
- Can I use this prompt if the two uploaded photos are of different people?Yes, you can use the same prompt structure for two different people. The original prompt specifies 'two versions of the same person' to create a mirror/doppelganger effect, but the AI will follow the reference images you provide. If you upload two distinct faces, the output will feature those two distinct faces interacting. This works very well for couple portraits or best-friend editorial shots. Just make sure both reference images are well-lit and facing relatively forward for the best connection with the straw.





