Load LoRA (LoraLoader)
Chains low-rank adaptation weights onto the UNet and CLIP streams with independent model and text-encoder multipliers. Package: Built-in Core.
Updated Oct 17, 2026 · 1 min read
- Node / Pack
- Load LoRA (LoraLoader)
- Category
- loaders
- Inputs
- model, clip, lora_name, strength_model, strength_clip
- Outputs
- MODEL, CLIP
- Source Pack
- Built-in Core
What Load LoRA (LoraLoader) does in a ComfyUI graph
Chains low-rank adaptation weights onto the UNet and CLIP streams with independent model and text-encoder multipliers.
On the canvas, this node accepts `model, clip, lora_name, strength_model, strength_clip` and emits `MODEL, CLIP`. Every pin is color-coded by data type so you cannot accidentally wire a pixel `IMAGE` tensor into a `LATENT` pin without an intermediate encode step.
Installation and missing-node recovery on Windows
Load LoRA (LoraLoader) ships inside the default ComfyUI core on Windows. Double-click anywhere on the empty Studio canvas and type its name to drop it into your graph.
Download the full Windows desktop or portable package at /download to get started with a clean Python 3.12 environment.
FAQs
How do I fix "Load LoRA (LoraLoader)" showing up as a red box in ComfyUI?
Update your ComfyUI installation to the latest build from /download so all recent core nodes are registered.
What are the main inputs and outputs of Load LoRA (LoraLoader)?
Inputs: model, clip, lora_name, strength_model, strength_clip. Outputs: MODEL, CLIP.
Run on your hardware. Free forever.
The full ComfyUI engine: open source, fast, extensible, and yours to run however you want. Download the desktop app and open this workflow on your own machine.
