ComfyUI Node
WAN Loop Frames
A ComfyUI node in conditioning/video_models with 5 inputs and 6 outputs.
WAN Loop Frames
- frames
- frames
- mask
- frame_count
- metadata
- context_frames
- raw_frames
◄frames_to_cut0►
◄frames_to_add0►
◄context_frames4►
◄preferenceadd frames►
Categoryconditioning/video_models
Inputs (5)
| Name | Type | Default | Description |
|---|---|---|---|
| frames | IMAGE | Full source animation. Only a midpoint context segment is output for sampling. | |
| frames_to_cut | INT | 00–9998 | Source frames removed alternately from the beginning and end; at least one frame is always preserved. |
| frames_to_add | INT | 00–9998 | Requested number of blank masked frames inserted at the midpoint seam. |
| context_frames | INT | 40–9998 | Number of source context frames included before and after the blank inpaint frames. |
| preference | COMBO | add frames | If the requested full-animation result is not 1 + 4n, round the blank-frame count up or down. |
Outputs (6)
| Name | Type | Description |
|---|---|---|
| frames | IMAGE | Only the requested context segment: before-context, blank frames, after-context. |
| mask | MASK | Black for context frames and white for the blank frames to inpaint. |
| frame_count | INT | Final full-animation frame count after WAN 1 + 4n adjustment. |
| metadata | STRING | Metadata for WAN Unloop Frames to merge the inpainted segment back into the full animation. |
| context_frames | INT | Actual number of frames in the context segment for the sampler length. |
| raw_frames | IMAGE | Untouched original frame batch for WAN Unloop Frames. |