Zombie AI Trend Prompt.
Choose your characters and copy an English storyboard for a video model that accepts two reference images. The template times recognition, approach, an embrace, and a memory across 15 seconds. Our generator already supplies the story, so no prompt is required there.
Explore the promptsOriginal English storyboard prompt
Zombie AI Trend prompt (optional)
Prefer to generate here? The studio already includes the story. Use this prompt with the matching reference video and two identity photos in a model that supports video references.
Zombie Apocalypse AI Video Editing Prompt
Zombie AI Trend prompt (optional)
Prefer to generate here? The studio already includes the story. Use this prompt with the matching reference video and two identity photos in a model that supports video references.
Official resources
ByteDance Seed · Seedance 2.5 ↗Set the reference photos in the right order
These prompts use Photo 1 for the survivor and Photo 2 for the person or pet who turns. In an external tool, attach two normal photos in that order and keep the Photo 1 / Photo 2 role instructions.
Our upload interface is arranged by role: the character who turns is on the left and the survivor on the right. Our generator automatically sends the survivor as the first model reference.
What you can change
Replace the ruined city with a quiet room or abandoned street, and the park memory with a beach or garden. For dogs and cats, add breed, coat, eye color, and distinctive markings.
Keep the identity, natural movement, and cut at the embrace. If your tool does not support 15 seconds or two reference photos, adapt the timing and input instructions to its actual capabilities.
When roles swap or faces drift
Check reference order and use sharp photos with one clear character per image. Avoid group photos, extra characters, and faces hidden behind objects.
Identity instructions cannot guarantee that a model follows them. For pets, preserve ears, coat, and markings, use lighter transformation details, and avoid human-like limb movements.
When the memory cut looks disconnected
Match character positions, camera angle, shoulders and arm placement, and movement across the cut. Reduce the number of shots and avoid many locations in one short clip.
If one clip cannot achieve the transition, generate the encounter and memory separately and edit them at first contact. Matching poses make alignment easier; still inspect the actual output.
Seedance 2.5 zombie AI trend prompt
Use the original 15-second brief below only in a workflow that accepts the intended reference images. Assign survivor to reference 1 and transformed partner to reference 2; adapt syntax and duration to your host. ByteDance’s model announcement describes broader reference and editing capabilities, but those are not the limits of this website: our current studio uses two portraits, a preset story and the duration shown in the generator. It supplies its own prompt. Check the actual charge before any external render.
15-second fictional reunion, two adult partners. Reference image 1: survivor. Reference image 2: transformed partner. 0–5s: a quiet encounter, recognition and lowered defenses. 5–10s: one careful step and a gentle embrace, natural hands. 10–15s: at contact, transition to a warm memory; keep both identities, shoulder positions, camera angle and motion direction. Subtle pale makeup in the encounter, normal appearance in the memory. No additional characters, attack or text.
Use a timed apocalypse storyboard
The copyable prompt proposes 15 seconds if the tool supports it: establish the encounter, recognize the person, lower defenses, approach, make contact, then continue in a healthy memory. When a model only makes shorter clips, divide the storyboard into encounter and memory shots rather than asking it to ignore its limit.
Describe the contact frame as an editing anchor
Specify the same camera distance, screen sides, shoulder positions, and movement direction at the transition. The warm shot should continue the embrace, with the transformed character healthy. Keep the environment change clear while the body movement stays connected; this makes the intended edit easy to describe and inspect.
Common questions
Are these prompts free to copy?
Yes. Reading and copying is free. External tools set their own generation prices, input requirements, and supported durations.
Do I need to paste a prompt into this site?
No. Our generator builds the story from your photos, relationship, and selected setting. These templates are for use in other tools that support reference images.
Why are the prompts in English?
We provide one English template for copying into different video tools. The instructions are available in English and Chinese. Check your external tool’s documentation for its supported prompt languages.