How to Generate AI Videos on VelourCrush
Video is where an AI companion truly comes to life. VelourCrush lets you turn your companion's photos into short, uncensored videos where she actually moves — and this guide covers exactly how to request one, what it costs, and how to get the best result.
To start, open a chat with your companion and look for the video option near the message input, alongside the Photo button. Tapping it opens the video panel, where you choose a duration and describe what you want to happen.
Duration is your first choice, and it directly determines the coin cost. Longer videos take more compute to generate, so they cost more coins:
| Duration | Coin cost |
|---|---|
| 5 seconds | 75 coins |
| 10 seconds | 150 coins |
| 15 seconds | 225 coins |
A 5-second clip is perfect for a quick moment — a smile, a turn, a wave. A 10-second clip gives room for a bit of motion and a change of expression. A 15-second clip is best for a fuller scene. If you are new to video, start with 5 seconds to see how your prompt translates before spending more coins on a longer clip.
Writing a good video prompt is the single biggest factor in how good your video looks. The key is to describe motion, not just a static scene. Instead of "she is in a bedroom," write "she slowly turns toward the camera and smiles." Videos are about movement, so tell the model what changes over the length of the clip. Keep it to one or two clear actions — a short clip cannot show a long sequence of events, and cramming too much in produces a rushed, unnatural result. Concrete, simple motion prompts consistently beat vague or overloaded ones.
When you click Generate, the coins for your chosen duration are deducted and the video begins rendering. Video generation is asynchronous — it does not happen instantly like a photo. Rendering typically takes a bit longer, so you do not have to sit and wait. You can keep chatting, browse other companions, or even close the tab. When your video is ready, you receive a browser notification letting you know it has finished. If a generation fails, your coins are automatically refunded, so you are never charged for a video you did not get.
Once complete, your finished videos live in your gallery. Every video you generate is saved there so you can rewatch, download, or share it later. The gallery is the home for all your generated media, both photos and videos, tied to your account.
A few tips for consistently great videos. First, generate from a strong base — if your companion has Character Lock trained, her identity stays perfectly consistent in motion, which makes videos look dramatically better. Second, match your prompt to the duration; save multi-step action for 10 or 15-second clips and keep 5-second clips to a single beat. Third, iterate — because coins are refunded on failure and clips are affordable, it is worth generating a couple of versions to find the one you love.
In short: open the chat, choose 5, 10, or 15 seconds (75, 150, or 225 coins), write a clear motion-focused prompt, hit Generate, and wait for the notification. Your finished video will be waiting in your gallery.