ComfyUI Node
✦ ADVANCED LORA LOADER
Stack any number of LoRAs in one node: on/off, strength, optional CLIP strength and a separate audio strength for audio-video models (LTX-2.3, MiniMax H3). Search picker with previews, Civitai info and trigger words; trigger_words carries the picked words.
✦ ADVANCED LORA LOADER
- model
- clip
- MODEL
- CLIP
- trigger_words
- lora_tags
◄stack{"v":1,"cache":false,"rows":[]}►
Categorymodel/loaders
Inputs (3)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | Diffusion model the LoRA stack is applied to. | |
| stack | STRING | {"v":1,"cache":false,"rows":[]} | The LoRA list, edited in the node UI (JSON). |
| clipopt | CLIP | Optional. Wire it when your LoRAs also train the text encoder (SDXL/Pony/SD1.5). |
Outputs (4)
| Name | Type | Description |
|---|---|---|
| MODEL | MODEL | — |
| CLIP | CLIP | Patched CLIP (empty when no CLIP is wired). |
| trigger_words | STRING | Trigger words picked in each LoRA's ⓘ panel, for the active LoRAs, comma separated. |
| lora_tags | STRING | <lora:name:strength> for each active LoRA — for metadata, captions or file names. |