# Raindance AI model and workflows Canonical: https://raindanceai.xyz/models.md Current public template model: **Seedance 2.0** (public model identifier: seedance-2). This is an independent site, not the model vendor's official service. Hidden legacy model catalogs and administrator-only experiments are not public offerings. Workflow: reference-photo video replacement using the selected existing performance video. The template supplies scene, motion and timing. Head swap and Whole person are two replacement instructions, not separate models. One photo for solo; two assigned photos for duet. Upload a clear head photo with full hairstyle for Head swap, or a full-body reference with face and clothing visible for Whole person. The photo preparer accepts readable raster photos up to 40 MB and rejects SVG/unreadable files; JPG/PNG exports are recommended. Output: one 720p MP4 per take. Fixed templates use portrait 9:16; the official source uses landscape 16:9 with preserved side padding. Official clip selection offers 4โ€“10 seconds, default 10 seconds, from the selected 30-second source. Fixed-template durations follow the catalog below. Original template audio is retained in delivery; no independent voice cloning or prompt-only generation is offered by this creator. Playback/appearance and content-review success are not guaranteed for every input. | Template | Public ID | Photos | Fixed duration / nominal duration (seconds) | | --- | --- | ---: | ---: | | Official Raindance ยท pick a 4โ€“10s moment | official-raindance | 2 | 8 | | Golden-Hour Pier | solo-golden | 1 | 8 | | Golden-Hour Duet | duet-golden | 2 | 8 | | Rain on the Pier | solo-rain | 1 | 8 | | Reception Duet | duet-reception | 2 | 8 | | Sunset Pier | pier-edge | 1 | 10 | | Summer Rain | rain-pier-man | 1 | 10 | | Beach & Cat | grandma-cat-sand | 1 | 10 | | Dock Close-Up | dock-student | 1 | 10 | | Sunset in the Sea | sea-wade | 1 | 10 | | Blue Hour & Lamplight | lamp-rain-man | 1 | 10 | The official template's nominal duration is catalog metadata; the creator's default selection is 10 seconds and users choose 4โ€“10 seconds. A full 30-second trend can be made as three separately charged 10-second takes and joined in an external editor. There is no separate stitching, training, image or layer-decomposition product on this site. See [pricing and credit rules](https://raindanceai.xyz/pricing.md) and the [creator](https://raindanceai.xyz/#create).