AI Zombie Trend
The AI Zombie Trend has taken social media by storm, letting anyone recast a dramatic short film where one partner turns into a zombie and the other must decide whether to pull the trigger. By uploading just two photos, users can insert themselves into a 24-second cinematic love story that cuts between a tense cabin standoff and golden-field memories. This page explains how the trend works, what tools are available, and what to expect from the generated video.
Official documentation
AI Zombie Trend: Event Overview
The AI Zombie Trend is a viral short-form video effect that reenacts a specific zombie love story scene. Users provide two photos — one for the survivor holding the gun and one for the person who becomes infected — and AI places those faces into every shot of the original clip, preserving the original timing, camera work, and audio.
What the Trend Looks Like
The source clip runs about 24 seconds in portrait 2:3 format. It opens in a dark, abandoned house where a man aims a gun at a woman who has turned into a zombie. He lowers the weapon and opens his arms, and the scene cuts to warm memories of the couple laughing, running, and sitting together in a golden field at sunset. The infected woman appears with milky eyes, cracked skin, and a snarl, while the healthy version of her appears only in the memory shots. The AI Zombie Trend preserves every cut, camera angle, lighting setup, film grain, and audio beat from the original, replacing only the people with the uploaded faces. The result is ready to download and post directly to TikTok, Instagram Reels, or YouTube Shorts without any additional editing.
How the Two-Photo System Works
Users upload exactly two clear, front-facing photos: Image 1 for the person holding the gun (the survivor) and Image 2 for the person who turns into a zombie. The AI first transforms the second photo into an infected, zombie-like portrait using a dedicated step — Nano Banana 2 on the imageat platform — that adds horror-film makeup while preserving the person's face shape, eyes, nose, lips, hairline, and hair. Then a motion-transfer model, Higgsfield Genjutsu on imageat, replaces the people in every shot: the survivor in all scenes, the infected portrait in the house scenes, and the healthy photo in the memory scenes. The workflow explicitly instructs the model never to mix the healthy and infected faces in the same shot. On Aura AI, a similar agent-based approach re-shoots all 16 shots of the source with the two faces and reassembles them at the original timing.
AI Zombie Trend: Confirmed Developments
Multiple platforms now offer dedicated tools for the AI Zombie Trend, each with confirmed workflows, pricing, and output specifications. The following details are verified from the source pages and reflect what users can expect when generating their own version.
