ImageOnlyCheckpointSave - ComfyUI Built-in Node Documentation

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

The ImageOnlyCheckpointSave node saves a checkpoint file containing a model, CLIP vision encoder, and VAE. It creates a safetensors file with the specified filename prefix and stores it in the output directory. This node is specifically designed for saving image-related model components together in a single checkpoint file.

Inputs

Parameter	Data Type	Required	Range	Description
`model`	MODEL	Yes	-	The model to be saved in the checkpoint
`clip_vision`	CLIP_VISION	Yes	-	The CLIP vision encoder to be saved in the checkpoint
`vae`	VAE	Yes	-	The VAE (Variational Autoencoder) to be saved in the checkpoint
`filename_prefix`	STRING	Yes	-	The prefix for the output filename (default: “checkpoints/ComfyUI”)
`prompt`	PROMPT	No	-	Hidden parameter for workflow prompt data
`extra_pnginfo`	EXTRA_PNGINFO	No	-	Hidden parameter for additional PNG metadata

Outputs

Output Name	Data Type	Description
-	-	This node does not return any outputs

CLIPSubtract - ComfyUI Built-in Node Documentation

LoraSave - ComfyUI Built-in Node Documentation

​Inputs

​Outputs

Inputs

Outputs