← Raindance AI

Raindance AI prompt guide

You do not need to write a prompt or adjust generation settings. Upload two portraits, choose who sings, and the service applies its fixed scene edit.

How the performers are mapped

  1. Scene: an approximately 15-second reference is used for an approximately 28-second edit with a 4:3 main frame, daylight, blue-green water, a wooden pier, and no other people in the background.
  2. Selection: upload two portraits, then choose which person sings.
  3. Mapping: the singer you select maps to the second image role used by the server; the other person maps to the pier role and its corresponding shots.
  4. Review: AI output can differ from the source portraits. Preview the edit and check identity, actions, and framing.

Google sign-in and purchase

The optional Google sign-in prompt in the home page header can be dismissed, and browsing needs no account. Click to make a video, choose a package, then use Google sign-in on the [pricing page](/pricing). After the callback, the selected package is preserved and PayPal checkout resumes automatically. The main flow needs no password or separate email verification. Confirm the offer and amount shown before payment.

Raindance AI is still in launch preparation. This page describes the current product flow; it does not mean generation results or soundtrack permissions have completed acceptance. Confirm you may use portraits before uploading. The service does not guarantee identical identity across frames or include music with cleared rights.

Create your video ↗