Nodes/sweet-tea-nodes/Wan VACE → Video (hybrid subject swap)
ComfyUI Node

Wan VACE → Video (hybrid subject swap)

Depth-first WAN VACE hybrid subject-swap node. Uses structural control video, a character reference, and optional source-RGB hybrid contexts for luma, detail, and background preservation.

By tea-time-labs·Created 6 months ago·Updated 23 days ago· 0
Wan VACE → Video (hybrid subject swap)
  • positive
  • negative
  • vae
  • control_video
  • source_video
  • control_masks
  • reference_image
  • positive
  • negative
  • latent
  • trim_latent
width832
height480
length81
batch_size1
strength1.00
control_strength-1.00
reference_strength-1.00
source_luma_strength0.25
source_detail_strength0.10
background_preserve_strength0.80
mask_feather_px8.0
Categoryconditioning/video_models

Inputs (18)

NameTypeDefaultDescription
positiveCONDITIONING
negativeCONDITIONING
vaeVAE
widthINT83216–16384
heightINT48016–16384
lengthINT811–16384
batch_sizeINT11–4096
strengthFLOAT1.000–4
control_strengthFLOAT-1.00-1–4
reference_strengthFLOAT-1.00-1–4
source_luma_strengthFLOAT0.250–4
source_detail_strengthFLOAT0.100–4
background_preserve_strengthFLOAT0.800–4
mask_feather_pxFLOAT8.00–128
control_videooptIMAGE
source_videooptIMAGE
control_masksoptMASK
reference_imageoptIMAGE

Outputs (4)

NameTypeDescription
positiveCONDITIONING
negativeCONDITIONING
latentLATENT
trim_latentINT