The AI zombie couple trend, made from your two photos

How to make a zombie couple video yourself

You do not need this page to make one. This is how the video is put together, with the prompts for every shot, so you can build it with any picture model and image-to-video model you like.

1. Step 1

   Draw one still per scene from your two photos

   Use a picture model that takes reference photos and draw four stills, one for each scene, from the same two photos. Drawing every still from the photos is what keeps the same two faces from the first shot to the last: ask a video model to invent them four times and you get four different couples.

2. Step 2

   Turn each still into a few seconds of video

   Give each still to an image-to-video model with one line about what moves in it: the stagger, the shaking hands, the sway, the spin. Ask for one continuous shot with no cuts, so every clip stays in its own scene and the cuts are yours to make.

3. Step 3

   Cut the four into one vertical video

   Lay the clips end to end at 9:16 and trim each to four or five seconds. Grade the street cold and the memory warm, put a flash of white before the memory, add one line of caption per scene, and use music you have the right to use, or none.

The shots, second by second

The prompts, shot by shot

Each shot is made in two steps. Give its still prompt to a picture model with the photos listed for that shot as references, then give the picture it draws to an image-to-video model with the motion prompt. Add these endings to every prompt: they set the look and keep the faces the same from shot to shot.

1. 0–4.5sNow

   Photo to attachWho turned

   Still prompt

   Copy

   A deserted city street at dusk after an outbreak: abandoned cars, scattered papers, light fog, overcast cold blue-grey light. In the middle of the street stands the person in the reference photo, turned into a zombie in a tender, gentle way: very pale grey-tinged skin, faint dark veins at the temples, clouded pale eyes, messy hair, their own clothes from the photo now dusty and torn. No blood and no wounds. Medium-wide shot, they face the camera with their head slightly tilted, lost and sad.

   Motion prompt

   Copy

   The zombie slowly staggers forward toward the camera, head tilting, arms hanging loose. Fog drifts across the empty street. Slow handheld push-in. Melancholic and cinematic, no blood.

2. 4.5–8.5sThe gun

   Photo to attachWho holds the gun

   Still prompt

   Copy

   The same deserted street at dusk, cold blue-grey light, background softly blurred. Close-up of the person in the reference photo gripping a handgun with both hands, arms extended straight forward, aiming at someone just off camera; the gun points ahead past the lens, never upward. Their eyes are wet with tears and their jaw is clenched; a smudge of dust on one cheek, their own clothes from the photo under a worn jacket. No blood.

   Motion prompt

   Copy

   Their hands tremble and the gun shakes. A tear runs down their cheek; they breathe hard and blink, unable to pull the trigger. Very slow push-in on their face. The gun is not fired: no muzzle flash, no shot.

3. 8.5–13sThe hug

   Photos to attach, in orderImage 1: Who turnedImage 2: Who holds the gun

   Still prompt

   Copy

   The same deserted street at dusk, cold light with a faint warm glow from a streetlamp. The person in image 2 holds the zombie version of the person in image 1 in a tight embrace; a handgun lies on the asphalt at their feet. The zombie (pale grey skin, clouded eyes, dusty torn clothes, no blood) rests their head on the other one’s shoulder, eyes half closed, calm. Both faces visible. Medium shot.

   Motion prompt

   Copy

   They hold each other closer and sway gently. The zombie's hand slowly relaxes and comes to rest on the other one's back. Fog drifts past as the camera slowly circles them. Tender and quiet, no blood.

4. 13–18sBefore

   Photos to attach, in orderImage 1: Who turnedImage 2: Who holds the gun

   Still prompt

   Copy

   A memory: a grassy field of wildflowers at golden hour, the low sun flaring behind them. The person in image 1 and the person in image 2, both healthy and happy in their own everyday clothes from the photos, laughing with their foreheads close together. Warm golden light, soft lens flare, gentle film grain.

   Motion prompt

   Copy

   They laugh and spin slowly together, hair moving in the breeze, the sun flaring behind them. Dreamy slow motion, gentle handheld camera. Warm and joyful.

Where it goes wrong

• Say who is who, shot by shot

  Given both photos at once, a picture model swaps the two people, most of all when they look alike, and the one who turned ends up holding the gun. Give a shot with one person in it only that person's photo, and say each person keeps the gender and age of their photo.

• Keep it a love story

  Ask for pale skin and clouded eyes, not blood or wounds, and say outright that the gun is never fired. Gore gets the request refused by the model or the post taken down, and it is not what the trend is about.

• Use clear photos, one person each

  Face fully visible, in good light, looking roughly at the camera. Sunglasses, heavy filters, group photos and faces taken from far away leave the model too little to keep each person recognisable.

• Only people who said yes

  Do not use celebrities or anyone who has not agreed to be in the video. A realistic video of a real person made without their consent can hurt them, and platforms take it down.

• Label it as AI-generated

  Switch on TikTok's AI-generated content label before you post. Realistic AI video of people is expected to carry it, and viewers trust a post that says so.