# Angry dance video generation and media

- Input: one photo of one adult (the user, or an adult who agreed). JPG, PNG or WebP, up to 10 MB. Clear front-facing face, ideally waist-up or head-to-knee, even light. Pets, cartoons and group photos are not supported.
- What changes: face, hairstyle, hair color, skin tone, build, clothing and accessories all come from the photo.
- What is fixed: the white kitchen, every gesture and its timing, the camera moves and the original sound. The user writes no prompt and cannot choose another dance, background or song.
- Routine (approximate): 0–1 s walk-in; 1–4 s finger pokes into the palm; 4–5 s open palm pushed to the lens; 5–10 s pointing, open hands and a shout; 10–12 s step back, hand on hip, turn aside; 12–14 s charge back for a final shout.
- Model: NanoAPI genjutsu-1.0 (motion control).
- Paid output: about 14 seconds, vertical 9:16, 720p MP4 with the original sound, no watermark. The home-page example is a real result generated at 480p.
- Sound: belongs to its rights holders and is not licensed by this site; use licensed music for ads or brand accounts.
- The instructions ask the model to keep the person's face, hair, skin tone and build, including when the hand covers the lens and during the side turn. Exact likeness is not guaranteed; check faces, hands and clothing.
- USD packs: 1 video $9.99, 3 videos $24.99, 10 videos $69.99; no subscription or free personal preview.
- Site photos: up to 7 days. Browser drafts: 24 hours. Finished videos: 30 days. Email links: 30 days. Unused credits do not expire.
- Confirmed generation failures return the attempt's credit once; this is not an automatic refund.

Independent service, not affiliated with TikTok, CapCut or ByteDance. Content is never published automatically.

Source: https://angrydancefilterai.com/
Support: support@angrydancefilterai.com