Guide to image and video generation

Upload a photo of a person, choose a mode and start the generation; the result appears in your Collection. Tokens are charged when you start, and if a generation fails for a technical reason they are refunded automatically.

One clear photo of one adult: the face and figure are clearly visible, without heavy blur, cropping or filters, in good light. JPG, PNG or WEBP up to 20 MB.

Only photos of people who are 18 or older and who have agreed to this kind of processing.

Photo modes change the picture following the mode's scenario. Pick a mode — its preview shows what you will get; the result is an image.

Video modes bring the photo to life with a short animation. Long videos are template-based clips of 15–35 seconds.

Video takes longer to generate: you can close the page — we will notify you when the result is ready.

Describe in words what you want: the outfit, the look, the details. Be specific and brief (up to 300 characters): “a fitted black evening dress with an open back” works better than “something beautiful”.

Requests pass a filter: attempts to bypass the rules, commands to the model and any mention of minors are rejected — no tokens are charged in that case.

Try another photo (better light, facing the camera), another mode, or rephrase the request. Generation is probabilistic: the same request can give different results.

If a generation fails, the tokens are refunded. If you still have a question, use “Contact support” in the result view.

Do not upload photos of minors or of people who have not given their consent, and do not use results for harassment, blackmail, deception, passing them off as real photos or sharing them without the depicted person's consent. See the Ethical Use Rules for details. Violations lead to account suspension.

Results are kept in your Collection: free ones for 30 days, paid ones for a year. You can delete any result from its view — its files will be erased.