In partnership with
LUXE PROMPTING
LUXE TREND WORKSHOP
3 Viral One-Photo Prompts, Rebuilt to Keep Your Face
This month's feeds are full of one-photo transformations: ten versions of you in a single carousel, yourself sealed inside a collector's box, a Halloween photo-booth strip. The ideas are fun. The common failure is not. Somewhere around panel four, the face stops being yours.
Below are three complete prompts rebuilt around one rule: lock identity first, then change everything else. Each comes with an example we generated for this issue so you can see what the wording does.
Last Chance to Invest in a Fast-Growing Popcorn Company
Popsmith’s investment round closes in a few weeks.
The brand is bringing popcorn back to the stovetop, with a patent-pending Popper and a growing line of kernels, oils, and seasonings.
Over 85,000 customers, $20M+ in lifetime revenue, and retail partners like Williams Sonoma, Costco, and Crate & Barrel show how far this premium popcorn brand has come in the $21B+ global market.
With revenue projected to jump more than 70% in 2026, the company is giving individual investors a chance to become shareholders. Get in before the round closes.
Explore the investment opportunity
The Short Version
- Three copy-paste prompts for this month's most-shared photo trends: the multiverse grid, the boxed action figure and the Halloween photo-booth strip.
- Each one names what must stay fixed (face, pose, light) before it names what changes.
- Swap in your own photo, run the prompt as written, then change one detail at a time.
How to use these prompts
Upload one clear, front-facing photo with even light and nothing covering your face. Paste the full prompt exactly as written into an image tool that accepts a reference photo, such as ChatGPT, Gemini or Grok Imagine. Replace anything in square brackets.
Our examples were generated from close versions of these prompts, with a fictional model described in place of a reference photo, so they show the structure of each prompt rather than how faithfully any one tool copies a real face. Use a photo of yourself, or one you have clear permission to use.
1. The multiverse grid: one face, six lives
The trend in its viral form asks for ten panels. Six is easier to keep consistent, and it still reads as a carousel when you crop each panel out. The trick is to fix the face, pose, camera height and background in one sentence before listing a single outfit.
The exact prompt
Original Luxe AI-generated example, created for this issue from a close version of the prompt below, with a fictional model described in place of the reference photo.
Using the uploaded photo as the only identity reference, create a 6-panel grid, 2 rows of 3, in a vertical 4:5 frame. Every panel keeps the exact same face, facial features, skin tone and hairstyle base as the reference, the same pose (standing, facing the camera, relaxed slight smile) and the same background: a city park in warm golden-hour light. Only the outfit and props change. Panels, left to right, top row then bottom: 1. astronaut in a plain white suit, helmet under one arm; 2. 1970s detective in a beige trench coat; 3. pastry chef in whites, holding a whisk; 4. jazz singer in a dark velvet jacket with a vintage microphone; 5. botanist in a field vest holding a fern; 6. cowboy in a brown hat and denim shirt. Photorealistic, consistent lighting and camera height across all panels, thin white gutters between panels. No text, no logos, no watermark.
What the result shows
The face, curly hairstyle, pose and golden-hour park hold steady across all six panels, which is the whole point of the identity sentence. One drift worth noticing: the cowboy panel added a bandana and work gloves the brief never asked for. Extra props tend to creep in on the last items of a long list.
One variation to try
Swap the six personas for six versions of your actual job across history: a 1920s, 1950s, 1980s, 2000s, today and 2080s edition. Keep the identity sentence word for word. If one panel's face drifts, regenerate the whole grid rather than patching a single panel.
2. The collector's box: your name in print
Newer image models handle short lines of text far better than they used to, and boxed action figures with a printed name are one of this month's most-shared formats. Quote the exact words, say where they go, and ban all other text.
The exact prompt
Original Luxe AI-generated example, created for this issue from a close version of the prompt below, with a fictional model described in place of the reference photo.
Turn the person in the uploaded photo into a collectible action figure, photographed inside its retail blister pack on a plain studio table. Keep the face exactly the same as the reference, sculpted in a slightly glossy vinyl finish at roughly 1:12 scale. The figure stands in a moulded clear plastic tray against a cardboard backing in deep plum and gold. To the right of the figure, three accessories sit in their own moulded slots: [ACCESSORY 1], [ACCESSORY 2] and [ACCESSORY 3]. Print the name "[YOUR NAME]" across the top of the box in bold condensed capitals, with the line "[YOUR TITLE] EDITION" beneath it in smaller type. Front-facing product shot, soft even softbox lighting, crisp reflections on the plastic, shallow depth of field on the table edge. Vertical 4:5. Spell all text exactly as written. No other text, no brand logos, no watermark.
What the result shows
Both text lines came back spelled exactly as quoted, and each of the three accessories sits in its own moulded slot. The model chose the figure's outfit on its own, a plum jacket that matches the packaging. Add a clothing line to the prompt if you want the figure dressed as you are in your photo.
One variation to try
Pick accessories that say something specific about you; our example used a sketchbook, a coffee cup and a camera. If the lettering comes back misspelled, shorten the title line before you change anything else.
3. The Halloween photo booth: four frames, one story
A photo-booth strip is a tiny storyboard. Give each frame one action and let the last one land the joke. Black-and-white film and an uneven flash do a lot of the vintage work for you.
The exact prompt
Original Luxe AI-generated example, created for this issue from a close version of the prompt below, with a fictional model described in place of the reference photo.
Create a vintage photo-booth strip of four stacked frames featuring the person in the uploaded photo. Keep the face, features and hairstyle exactly the same in every frame. Set it in October: a dark velvet booth curtain behind them, a small carved pumpkin on the ledge. Frame 1: looking straight at the camera with a calm smile. Frame 2: holding a paper bat mask up to one eye. Frame 3: laughing, head tilted back. Frame 4: a quick startled glance over the shoulder, with only a faint pale shape in the curtain folds beside them. Black-and-white film look with soft grain, slightly uneven booth flash, thin white borders, and one hand-stamped date at the bottom reading "OCT 31". The strip lies on a plain dark wooden table, shot from above in a vertical 9:16 frame. No other text, no logos, no watermark.
What the result shows
All four actions landed in order, the pumpkin repeats in every frame and the OCT 31 stamp is legible. The ghost in frame 4 came out as a clear figure rather than the faint shape the brief asked for. If you want subtler, try "barely visible" and "half hidden in the curtain folds."
One variation to try
Change only frame 4. Replace the pale shape with a black cat walking along the ledge, or a friend leaning in from the edge of the curtain. Keep frames 1 to 3 identical so you can see how one change alters the story.
A quick identity check before you post
- Look at the eyes, nose and hairline first. If they drift in one panel or frame, regenerate the whole image.
- Zoom into hands and props. Accessories and microphones are where small errors hide.
- Read every word of printed text letter by letter.
- Label realistic AI images where the platform asks you to.
Make one of these yours
You do not need to try all three. Pick one, make a first image, then change a single instruction and compare the two.
When you share it with the community, include the result, the exact prompt, the tool or model and any settings you know. Add one sentence about what you changed and whether it helped.
Source Desk
All three prompts are original Luxe editorial work, written for this October 2026 issue and informed by trend coverage of one-photo AI transformations this month, including the multiverse persona carousel, boxed action figures and photo-booth strips. Example images were generated for this issue from fictional descriptions; no real person's photo was used. Result notes describe these particular images.
Multiverse persona trend (The Prompt Master) · October AI photo trends (Quantslant) · Luxe Prompting
A clearer brief. A more useful result.