Nodes/ComfyUI-SwissArmyKnife/VACE Annotator - Depth
ComfyUI Node

VACE Annotator - Depth

Loads a VACE depth checkpoint (MiDaS or Depth Anything v2), processes IMAGE batches at the chosen resolution, and returns depth maps for downstream control signals.

By sammykumar·Created 11 months ago·Updated 7 months ago· 5
VACE Annotator - Depth
  • images
  • depth_maps
model_typemidas
resolution512
model_path
CategorySwiss Army Knife 🔪/VACE Annotators

Inputs (4)

NameTypeDefaultDescription
imagesIMAGEInput images or video frames (batch supported)
model_typeCOMBOmidasDepth estimation model to use
resolutionINT51264–2048Processing resolution (lower = faster, higher = more detail)
model_pathoptSTRINGCustom path to model file (leave empty for default: models/vace_annotators/depth/)

Outputs (1)

NameTypeDescription
depth_mapsIMAGE