How to make an AI Raindance video
Learn how to make an AI Raindance video from photos: Google sign-in, solo or duo roles, video credits, original audio and a 15-second 768P MP4 download.
Start with a clear photo
Choose a well-lit JPEG, PNG or WebP portrait with one clearly visible face. Avoid sunglasses, heavy filters and crowded group photos. Use your own photo or a photo you have permission to use.
Choose solo or duo
Solo uses one photo for the singer and removes the companion. Duo uses two photos: the first is the companion and the second is the singer. Swap the roles before generating if needed.
Generate and come back later
Sign in with Google and use one video from your balance. The first template follows the opening 15 seconds of the reference with a 4:3 frame, 768P output and matching audio. Your task appears in My videos; you can return to it after refreshing.
Download and check the performance
Check faces, role assignment and singing timing before sharing. AI can vary from the reference; exact facial expressions and perfect lip sync are not guaranteed. Download within 30 days. A confirmed failed task returns its video credit once.