How to Generate AI Images on VelourCrush
Generating an AI image on VelourCrush takes about ten seconds of setup and one click. This guide walks you through the whole process, explains what every option does, and shares tips for getting the best possible photo of your companion.
Start by opening a chat with any companion. Every companion on the platform can generate images, whether she is one of our curated characters or one you built yourself in the creation wizard. Once you are in the chat, look for the Photo button near the message input. Tapping it opens the image panel.
The image panel is built around act tiles. Each tile describes what your companion is doing in the photo. Tap one to select it. Here is what the common tiles mean:
| Act tile | What you get |
|---|---|
| Portrait | A clean, clothed or lightly styled photo focused on her face and pose |
| Oral | An explicit oral scene |
| Riding | An explicit cowgirl-position scene |
| Doggy | An explicit rear-entry scene |
| Missionary | An explicit face-to-face scene |
| Lingerie | Her posing in lingerie, suggestive but not explicit |
You can pick the tile that matches the moment. If you only want a tasteful photo, choose Portrait or Lingerie. If you want something explicit, pick the act that fits. The tiles exist so you do not have to describe an entire scene in words — the tile handles the composition and the model does the rest.
Next, choose a setting. The setting controls where the photo takes place — bedroom, shower, beach, office, poolside, and more. The setting changes the lighting, background, and mood of the image, so pick one that fits the story you and your companion are in.
After the tile and setting, you can add a custom description. This is a free-text box where you fill in the details the tiles cannot — an outfit color, an expression, an accessory, a specific angle. Keep it short and concrete. "Biting her lip, wearing a red silk robe" works better than a long paragraph. The description layers on top of the act tile and setting rather than replacing them.
When everything looks right, click Generate. Each image costs 10 coins. The coins are deducted when generation starts, and if the generation fails for any reason, the coins are automatically refunded to your balance — you are never charged for a photo you did not receive. Generation usually completes in a few seconds, and the finished image appears directly in your chat and is saved to your gallery.
For the best results, keep three things in mind. First, let the act tile and setting do the heavy lifting and use the description only for fine detail. Second, be specific but not contradictory — asking for a beach setting and a bedroom detail in the description confuses the model. Third, generate a few variations; small changes in the description produce noticeably different photos, and coins are cheap enough to experiment.
One of the most powerful features in the image panel is Pose reference. Instead of leaving the pose to chance, you can upload a photo of a pose, or choose one from the built-in pose library, and the generated image will follow that pose. This is the easiest way to get a specific body position or framing without describing it in words. If you have a particular pose in mind, Pose reference gets you there far more reliably than text alone.
Put together, the workflow is simple: open a chat, tap Photo, pick an act tile, pick a setting, add a short description, optionally set a Pose reference, and hit Generate. Ten coins later you have a new photo of your companion — and if you want her to look identical in every single image, pair this with Character Lock for perfect consistency.