Swap People & Objects in Any Video
Motion-control style dual upload. Default Genjutsu Light for ≥1:4 margin; Full is optional premium.
- Source video + reference still (dual slots)
- Default Genjutsu Light supply path
- Required Credits before Generate — no free grant
Supply: Genjutsu Light UI ready; HF API OFF — CLI/web assets for beautify only. Full = premium tier.
Supply path: Genjutsu Light (Full optional via HD credits).
Upload the clip whose motion you want to keep
Face / product / object to insert
Default Genjutsu Light · 720p. HD uses higher credits.
Supply: Genjutsu Light · 720p · Required Credits lock: 80 cr Light / 200 cr Full path (≥1:4 vs ~$0.05/s Light).
Genjutsu Light supply · Object Swap / Motion Transfer.
after_n_runs · standard 80 cr · HD 200 crHow it works
- Upload motion — Add the source video (walk / gesture / performance).
- Upload subject — Add the reference image for the person or object to swap in.
- Describe the swap — Optional prompt — what should change vs stay.
- Generate & download — Pay Required Credits; unlock export after preview.
Features
Dual upload
Source motion video + subject reference — Motion Control skeleton, Genjutsu copy.
Light by default
Genjutsu Light keeps cost:credit ≥1:4; Full optional via HD credits.
Identity-aware swap
Replace cast or product while preserving timing.
Credits gate
Google sign-in + packs; no no-login free export.
Use cases
Cast swap
Put a new face on an existing performance clip.
Product insert
Drop a SKU into motion footage for ads.
UGC restyle
Reuse strong motion with a new subject.
FAQ
- Light vs Full?
- Blinkly defaults to Genjutsu Light. Full uses the HD credit tier.
- Keys ready?
- Page + creditCost ship now. HF API stays off; site uses static beautify assets only.
- vs talking photo?
- Talking photo is still→speech; Genjutsu is motion + subject swap.
Related tools
Margin target ≥1:4. Backend higgsfield/genjutsu-light.