How to cast your photos in an AI video template
Written by Kadenly · Updated
Choose a scene, upload one or two portraits and map the singer for a newly generated character-cast video.
Purpose and limitations
Create a new character-cast video from one or two photos. Choose a prepared MV or scene, assign the singer and partner, and receive music and silent exports. This is photo-based scene creation, not an editor for arbitrary source footage.
Choose a complete template
Select the 15-second Raindance MV or an 8-second vertical scene. The template sets the background, duration, format and music.
Upload your cast
The MV needs two individual photos. Other scenes support Solo or Duet. In Duet, choose which photo supplies the singer; the other person is the partner.
Generate, review and download
Check the displayed credit quote, then generate. Review both identities and all shots, listen to the music version and download it or the silent export.
English prompt example
This is the editable English preset from the tool: @Image1 and @Image2 map to the two characters; @Video1 is your uploaded reference clip. It guides new synthesis rather than frame-locked replacement.
Create a 15-second seaside music-video performance with two distinct adults. Use @Image1 for performer A and @Image2 for performer B. Keep each identity, hairstyle and visible outfit consistent across shared shots, solo shots and close-ups. Performer A is the singer and performer B is the partner. Use a prepared seaside setting rather than either portrait background. Keep faces readable, natural body proportions and coherent lighting. No extra people, merged faces, swapped identities, captions or logos. Generate silent visuals; the fixed soundtrack is composed afterward.Input and result checklist
- Choose the template before uploading. The MV uses two portraits; the other scenes offer Solo and Duet.
- Choose the singer by photo number, not left/right position: people can move between shots.
- Use clear individual portraits with visible outfits. Review identities after every cut, including solo close-ups.
- The cover is AI concept art, not a verified generated result. Shot timing and mouth movement are approximate.
Common issues and adjustments
What does generation cost?
The selected template has a fixed duration and format. Its current credit quote is shown before generation; no reference-video upload or reference-input processing charge is added. A recorded failed task restores deducted credits. A completed creative result remains charged.
How should I prepare the photos?
Use one clear adult portrait per slot, with the face unobstructed and the clothing you want visible. Separate individual portraits give clearer identity references than group photos.
Can I use it for a couple or friends?
Yes. Duet casts photo 1 and photo 2 as two distinct people. In single-shot presets, Solo needs one photo. The main MV template always uses two people.
Sources and related policies
These are original preparation tips and prompt examples from our team. Model capabilities change; check current settings before generating. For input requirements and usage rules, see: