ARC Power LoRA

ARC Power LoRA Loader (Model Only)

Minimal ARC-native ComfyUI node for stacking multiple LoRAs without any CLIP passthrough.

Node

ARC Power LoRA Loader (Model Only)

• Input: model
• Output: model
• 8 LoRA slots
• Each slot has a LoRA dropdown and model strength
• [None] disables a slot
• LoRAs are applied sequentially from slot 1 through slot 8
• No CLIP input, output, or patching

Install

Copy the ComfyUI-ARC-Power-LoRA folder into:

ComfyUI/custom_nodes/

Then restart ComfyUI.

The node appears under:

ARC > loaders

Workflow setup

Place it immediately after your diffusion model loader and before ARC's memory / attention / sampler path.

Example:

Load Diffusion Model -> ARC Power LoRA Loader -> Configure Model Memory -> Configure Attention Backend -> Sampler

Once the workflow is working, export the ComfyUI workflow/API JSON and send it back for ARC integration.

Notes

This first version intentionally uses fixed slots instead of a frontend JavaScript dependency. That keeps the node self-contained and stable while still supporting a LoRA stack. The slot count can be increased easily later.