AI Rap Flow · MUSIC TRENDS
Raindance AI
Two photos. One fixed waterfront performance. Keep your faces and outfits in a 27.04-second silent Raindance AI video.
Start creating ↗Your photos stay private · Review your settings and credits before creating
What is Raindance AI?
Raindance AI is a music-video trend associated with Raindance by Dave and Tems. The waterfront performance brings two people into a staged duet. Our template keeps its fixed choreography and uses two photos for the faces and outfits, with a silent 27.04-second delivery.
Raindance AI video ideas
Cast a Raindance AI video with two photos—one for each role in the seaside duet. The Raindance AI template keeps the pier cuts, movement, and music; your photos supply both performers. The clip is an AI-generated example from our studio, not the original music video or a template reference. Generate your own version with your photos.
Unable to play this example. Try again.
Seaside Raindance duet
Your Faces and Outfits. The Raindance AI Performance.
Two photos. One fixed waterfront performance. Keep your faces and outfits in a 27.04-second silent Raindance AI video.
How to Make Your Raindance AI Video
1. Upload two separate photos: Person A is the seated dock partner; Person B is the singer. Include a clear face and the outfit you want recreated. Roles do not depend on gender.
2. Keep Raindance AI selected and choose H3 768P for 40 credits or H3 2K for 65 credits. The 27.04-second duration, 4:3 framing and choreography are fixed.
3. Review your photos, selected quality, credit cost and tested three-part prompt before confirming. The finished delivery format is a silent MP4 saved to your account, ready for your own soundtrack.
Fixed Camera Cuts, Movement and Mouth Timing
Raindance AI uses the tested reference-video workflow. Your two photos define each person’s face, skin tone, hairstyle, age appearance and clothing. The reference defines the waterfront setting, seated poses, gestures, camera angles and shot-cut timing.
Three H3 segments are generated at 9, 9 and 10 seconds. FFmpeg removes the technical padding and joins the 8.72, 8.72 and 9.60-second action windows into a 27.04-second, 25fps video. Reference audio guides mouth timing; the delivered MP4 has no audio track. Preview images are labeled placeholders.