Skip to main content
The QwenImageDiffsynthControlnet node applies a diffusion synthesis control network patch to modify a base model’s behavior. It uses an image input and optional mask to guide the model’s generation process with adjustable strength, creating a patched model that incorporates the control network’s influence for more controlled image synthesis.

​
Inputs

ParameterDescriptionData TypeRequiredRange
modelThe base model to be patched with the control networkMODELYes-
model_patchThe control network patch model to apply to the base modelMODEL_PATCHYes-
vaeThe VAE (Variational Autoencoder) used in the diffusion processVAEYes-
imageThe input image used to guide the control network (only RGB channels are used)IMAGEYes-
strengthThe strength of the control network influence (default: 1.0)FLOATYes-10.0 to 10.0
maskOptional mask that defines areas where the control network should be applied (inverted internally)MASKNo-
Note: When a mask is provided, it is automatically inverted (1.0 - mask) and reshaped to match the expected dimensions for the control network processing. The node uses different internal processing methods depending on whether the model patch is a ZImage Control type or a standard DiffSynth control network.

​
Outputs

Output NameDescriptionData Type
modelThe modified model with the diffusion synthesis control network patch appliedMODEL
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub
Source fingerprint (SHA-256): 61833984d0b92be65fae72a894806572c0588dea74a295e8289d1194dee611bb