CheckpointLoader - 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 CheckpointLoader node loads a pre-trained model checkpoint along with its configuration file. It takes a configuration file and a checkpoint file as inputs and returns the loaded model components including the main model, CLIP model, and VAE model for use in the workflow.

Inputs

Parameter	Data Type	Input Type	Default	Range	Description
`config_name`	STRING	COMBO	-	Available config files	The configuration file that defines the model architecture and settings
`ckpt_name`	STRING	COMBO	-	Available checkpoint files	The checkpoint file containing the trained model weights and parameters

Note: This node requires both a configuration file and a checkpoint file to be selected. The configuration file must match the architecture of the checkpoint file being loaded.

Outputs

Output Name	Data Type	Description
`MODEL`	MODEL	The loaded main model component ready for inference
`CLIP`	CLIP	The loaded CLIP model component for text encoding
`VAE`	VAE	The loaded VAE model component for image encoding and decoding

Important Note: This node has been marked as deprecated and may be removed in future versions. Consider using alternative loading nodes for new workflows.

​Inputs

​Outputs

Inputs

Outputs