Nodes/ERPK Collection/Kling 3.0 Image-to-Video
ComfyUI Node

Kling 3.0 Image-to-Video

A ComfyUI node in ERPK/WaveSpeedAI with 15 inputs and 1 output.

By eRepublik-Labs·Created 10 months ago·Updated about a month ago· 0
Kling 3.0 Image-to-Video
  • image
  • end_image
  • client
  • video_url
modelKling 3.0
prompt
image_url
duration5
seed-1
negative_prompt
end_image_url
cfg_scale0.50
soundfalse
shot_typecustomize
multi_prompt
element_list
CategoryERPK/WaveSpeedAI

Inputs (15)

NameTypeDefaultDescription
modelCOMBOKling 3.0Model variant: Kling 3.0 (standard), Kling 3.0 Pro (higher quality), or Kling 3.0 4K (highest resolution; derives aspect ratio from the input image)
promptSTRINGText description of the desired motion/scene
image_urlSTRINGStarting image URL (use WaveSpeed Upload Image to produce one). Fallback when the IMAGE input is not connected.
imageoptIMAGEStarting image as a ComfyUI IMAGE tensor. Preferred — takes precedence over `image_url` when connected.
end_imageoptIMAGEOptional end-frame guidance image as a ComfyUI IMAGE tensor. Preferred — takes precedence over `end_image_url` when connected.
clientoptWAVESPEED_AI_API_CLIENTWaveSpeed API client (optional if API key is configured in Settings)
durationoptINT53–15Video duration in seconds (3 to 15)
seedoptINT-1-1–2147483647Random seed for reproducibility (-1 for random)
negative_promptoptSTRINGElements to exclude from the generation
end_image_urloptSTRINGEnd-frame guidance image URL. Fallback when the end_image IMAGE input is not connected.
cfg_scaleoptFLOAT0.500–1Prompt adherence strength, 0-1
soundoptBOOLEANfalseEnable synchronized audio generation (applies a 1.5x cost multiplier)
shot_typeoptCOMBOcustomizeShot composition mode: 'customize' or 'intelligent'
multi_promptoptSTRINGJSON array of scene-segmented prompts (mutually exclusive with prompt on 4K)
element_listoptSTRINGJSON array of pre-generated element IDs for visual consistency

Outputs (1)

NameTypeDescription
video_urlSTRING