EasyCache - ComfyUI Built-in Node Documentation

The EasyCache node adds a native caching system to a diffusion model that speeds up sampling by reusing results from previously computed steps instead of recomputing every step. It activates only between a configurable start and end point of the sampling process, and skips steps when the estimated output change stays below a user-defined threshold. This is an experimental node intended for advanced debugging use.

Inputs

Parameter	Description	Data Type	Required	Range
`model`	The model to add EasyCache to.	MODEL	Yes	-
`reuse_threshold`	The threshold for reusing cached steps (default: 0.2).	FLOAT	Yes	0.0 - 3.0
`start_percent`	The relative sampling step to begin use of EasyCache (default: 0.15).	FLOAT	Yes	0.0 - 1.0
`end_percent`	The relative sampling step to end use of EasyCache (default: 0.95).	FLOAT	Yes	0.0 - 1.0
`verbose`	Whether to log verbose information (default: False).	BOOLEAN	Yes	-

Outputs

Output Name	Description	Data Type
`model`	The model with EasyCache functionality added.	MODEL

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): 3e10ac65f8df58ce8649fdf599e62bfb86f2d4166840bed5622c0aa2c419cd38

​Inputs

​Outputs

Inputs

Outputs