Wan 2.1 trajectory-guided video
Produces a silent video from a starting image, a motion prompt and point trajectories using the fixed Wan 2.1 ATI model.
Loading this version’s controls and setup…
Workflow details
Workflow details
Runtime profile
- Publisher
- Sweet Tea Automation
- Status
- Listed
- Version
- v0.1.6
- Updated
- Nov 4, 2026
Hover nodes for Pipe controls
Core
Primary Pipe surface
Source image
LoadImage
Positive prompt
CLIPTextEncode
Motion trajectories
PrimitiveStringMultiline
Video shape
WanTrackToVideo
Extended
Additional surfaced controls
Negative prompt
CLIPTextEncode
Image guidance
CLIPVisionEncode
Sampling
KSampler
Noise schedule
ModelSamplingSD3
Video playback
CreateVideo
Video encoding
SweetTeaPreviewVideo
Requirements
Models
4Clip Vision · 1.18 GB · SAFETENSORS · FP16 + I64 (tensor header) · Hugging Face · Comfy-Org/Wan_2.1_ComfyUI_repackaged
Text encoder · 6.27 GB · SAFETENSORS · U8 + FP16 + FP32 + FP8 (tensor header) · Hugging Face · Comfy-Org/Wan_2.1_ComfyUI_repackaged
VAE · 242 MB · SAFETENSORS · BF16 (tensor header) · Hugging Face · Comfy-Org/Wan_2.1_ComfyUI_repackaged
Wan2_1-I2V-ATI-14B_fp8_e4m3fn.safetensors
Not resolvedDiffusion model · Hugging Face · Kijai/WanVideo_comfy
Node packs
2- Found
Node pack · Unknown
Node pack · GitHub · tea-time-labs/sweet-tea-nodes
Sources
1 source
Sources
1 sourceModel artifacts
1 artifact
Model artifacts
1 artifact66b0b12b-1ef4-4923-b3ec-e36f14e2a61b-comfy-video-wan-ati-draft-0.1.6.tea
tea · 25.9 KB · SHA-256 197eaf0ffedc…c19d
Download file