Nodes/HunyuanVideo-1.5 nodes/HunyuanVideo Prepare Latents
ComfyUI Node

HunyuanVideo Prepare Latents

A ComfyUI node in HunyuanVideoWrapper1.5 with 6 inputs and 4 outputs.

By yuanyuan-spec·Created 9 months ago·Updated 8 months ago· 27
HunyuanVideo Prepare Latents
  • hyvid_cfg
  • vae
  • latents
  • reference_image
  • latents_dict
  • height
  • width
  • n_tokens
aspect_ratio16:9
target_dtypebfloat16
CategoryHunyuanVideoWrapper1.5

Inputs (6)

NameTypeDefaultDescription
hyvid_cfgHYVID15CFG
vaeHYVID15VAE
aspect_ratioSTRING16:9
target_dtypeCOMBObfloat169 options: float32, float64, float16, bfloat16, uint8, int8, +3
latentsoptLATENT
reference_imageoptIMAGE

Outputs (4)

NameTypeDescription
latents_dictHYVID15LATENTSDICT
heightINT
widthINT
n_tokensINT