Most image-to-video demos still assume you will write a prompt and then clean up the result in an editor. That is a poor fit when you already know the scene you want and you only need a short MP4 for a chat or a post.
I have been looking at a narrower design: published video templates that already contain choreography, camera and sound. You supply the people as still photos. The product I used as a reference is AI4MEME, a browser-based studio with 6 scenes and no install.
What is fixed vs what you provide
Each scene is a real template, not a blank prompt. A card shows the photo count, length, aspect ratio and credit price before you start. Examples from the current library:
- No Scrubs: 3 photos, 16 seconds, 4:3, 1520 credits, up to 720p
- Infinity Pool: 1 photo, 11 seconds, 9:16, from 190 credits (480p) or 480 credits (720p)
- Zombie Love: 2 photos, 26 seconds, 9:16, from 820 credits
Your job is to add one clear adult photo per role (JPEG, PNG or WebP, up to 8 MB) and confirm consent. The scene owns wardrobe, motion and the original soundtrack. The site is explicit that AI-generated body shape may not match the photo exactly.
Formats and output
Outputs are scene-specific MP4 files in 4:3, 16:9 or 9:16, between 5 and 26 seconds. That covers a throwback dance break, a vertical story clip, and a short lifestyle shot without you picking a codec or a timeline.
Credits, not a mystery meter
Credits are deducted when you start the creation. Pack credits never expire; monthly plan credits expire at the end of the billing month and do not roll over. Unused pack credits stay if you cancel a plan. Failed renders refund automatically. Plan credits are used before pack credits.
You do not need a local GPU or an API key. It runs in the browser, including on a phone. Sign-in is with Google.
If you want the public scene list and current prices, they are on AI4MEME.
Top comments (0)