Wiki/Video/Video Generation

🎬Video Generation

How video works, the models, costs, and tips.


AI Peeps can turn an image into a short video — your Peep moving, dancing, kissing, even talking.

How it works

Every video starts from a first frame — an existing image of your Peep (or any image you've made on the site). You can optionally set a last frame to control where the motion ends. Then add an action and/or a short prompt.

Models: Base vs Plus

The model is your first choice — every other option adapts to it.

  • Base (default) — a solid balance of quality and cost; sticks closely to your prompt and supports the full feature set, including talking videos, end frames, and loops.
  • ✨ Plus (premium) — higher quality and noticeably more reliable motion, with far fewer of the weird, off-target results Base can sometimes produce. It also adds motion-intensity presets. Plus is image-to-video only — no talking videos, end frames, or loops — and costs about 2× Base.
❗ Important
Base and Plus have different action sets. When you switch model the list of available actions changes — a few actions exist on one tier only.

What it costs

Cost depends on resolution (480p / 720p / 1080p), length (5 / 10 / 15s), and the model. Rough guide for Base:

Resolution5s10s15s
480p4075110
720p100190270
1080p200380550

Plus costs roughly double. Looping, talking, enhancing and extending also cost more. The exact, live cost is always shown before you confirm — video is the most coin-intensive feature, so check it there. (And you're never charged for a video that fails — see Troubleshooting.)

Tips

  • Start small, then enhance. Generate a base video first; once you like it, Enhance it to higher resolution and 60 fps.
  • Test short. Make a 5- or 10-second clip first, and only Extend it if it turns out well.
  • The first frame matters most. Pick a starting image the motion can naturally flow from — see Video Actions.
  • NSFW = Action, not prompt. Always pick an NSFW action for explicit movement.

Keep exploring