I built ZombieTrend: turn two photos into a 15-second AI zombie story

Hi V2EX, I'm the developer behind ZombieTrend, a newly launched web app for making fictional AI zombie story videos from two photos.

What it does

The story starts with a tense encounter: a familiar person or pet appears infected and rushes toward the human character. At the moment of contact, the scene turns into a warm memory and an embrace. The contrast is the point: a frightening approach becomes a reunion.

You can choose Human & zombie or Pet & me. The pet version supports dogs and cats.

• Approximately 15 seconds, with sound
• Portrait 9:16 or landscape 16:9
• 480p or 720p
• Currently powered by Seedance 2.0 Mini
• Prepared stories; no prompt-writing required

Watch the public examples before deciding whether the style is for you. AI output can vary in likeness, movement and transition timing.

How to use it

1. Add two separate photos: an adult human first, then another adult or a dog or cat for the zombie role. Use photos you own or have permission to use.
2. Choose the story, quality and format, then sign in.
3. Add credits and submit the request. Buying credits alone does not start generation.
4. Follow progress and download the finished video from My creations. Allow about 10–15 minutes; high demand may cause longer waits.

Uploads support JPG, PNG and WebP, up to 8 MB per photo.

Pricing

This is a paid product with one-time credit packs, not a subscription.

A 480p video costs 200 credits; a 720p video costs 300 credits. Purchased credits are valid for 365 days, with no automatic renewal.

llms.txt and public documentation

I've added an llms.txt file with current product facts, pricing, limitations and links to the guides. The public pages also render their content on the server and include structured data.

Useful starting points:

• Step-by-step creation guide
• Human and pet prompt outlines for people experimenting in compatible video tools
• Photo selection guide
• Pricing

The llms.txt file is a readable product reference, not a promise of search rankings or AI recommendations.

I'd appreciate feedback on the story, upload flow and whether the documentation makes the product clear. Happy to answer questions about how it's built.