ChromaRadianceOptions - ComfyUI Built-in Node Documentation

The ChromaRadianceOptions node allows you to configure advanced settings for the Chroma Radiance model. It attaches a wrapper to an existing model and applies the selected options during the denoising process only when the current sigma value falls within the configured range, giving control over NeRF tile size and text token ID handling.

Inputs

Parameter	Description	Data Type	Required	Range
`model`	The model to apply Chroma Radiance options to	MODEL	Yes	-
`preserve_wrapper`	When enabled, will delegate to an existing model function wrapper if it exists. Generally should be left enabled. (default: True)	BOOLEAN	No	-
`start_sigma`	First sigma that these options will be in effect. (default: 1.0)	FLOAT	No	0.0 to 1.0
`end_sigma`	Last sigma that these options will be in effect. (default: 0.0)	FLOAT	No	0.0 to 1.0
`nerf_tile_size`	Allows overriding the default NeRF tile size. -1 means use the default (32). 0 means use non-tiling mode (may require a lot of VRAM). (default: -1)	INT	No	-1 and above
`force_sequential_txt_ids`	Force usage of sequential text token IDs instead of zeroes. Should be used for checkpoints from 2026-05-22 to 2026-06-01 that are trained in this way but do not contain the sequential key in the state dict. (default: False)	BOOLEAN	No	-

Note: The Chroma Radiance options only take effect when the current sigma value falls between end_sigma and start_sigma (inclusive). The nerf_tile_size option is only applied when set to 0 or a higher value (a value of -1 uses the default tile size of 32 and stores no override). The force_sequential_txt_ids option is only applied when set to True. When nerf_tile_size is -1 and force_sequential_txt_ids is False, no options are configured and the model is returned unchanged without any wrapper applied. Note: All inputs other than model are advanced options.

Outputs

Output Name	Description	Data Type
`model`	The model with Chroma Radiance options applied, or the unchanged model if no options are active	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): 761f1946fe1fd77158e97f6f34d002e2445cc00e008741f8c37cde5673900409

​Inputs

​Outputs

Inputs

Outputs