LoRA Loader Pixaroma

LoRA Loader Pixaroma is the pack's answer to the "eight LoraLoaders in a row" problem. Stack as many LoRAs as you want in one compact node, each on its own line with an on/off switch and a strength you type or nudge with arrows - and chain the model and clip through several of these nodes when one stack isn't enough. The LoRAs apply in row order, which matters because stacking LoRAs (a style plus a character plus a concept) is the standard workflow move, and their interactions are famously unpredictable.

The part that makes it more than a tidy LoraLoader: click the i on any row and the node reads the LoRA's info and trigger words straight from the file - no internet needed. Tick the ones you want (or type your own) and they come out of a triggers output as plain text you wire into your prompt. That kills the classic chore of keeping trigger words in a separate text file and remembering which LoRA they belong to.

Inputs and outputs

• model (required) - the diffusion model every switched-on LoRA is applied to.
• clip (optional, but recommended) - the CLIP text encoder the LoRAs also get applied to. Wire the checkpoint CLIP in here and the CLIP output on to your text encode, so LoRAs can tune how their trigger words are read, not just how the image is drawn. Leave it unwired only for a model-only setup.

Outputs:

• MODEL - the model with every switched-on LoRA applied, in row order.
• CLIP - the CLIP with every switched-on LoRA applied (passes through unchanged if no CLIP was connected).
• triggers (STRING) - the trigger words you picked, from switched-on LoRAs only, joined as plain text for your prompt.

Civitai lookup

There's an optional Civitai lookup that fetches official trigger words and a preview when you ask. If it keeps saying it can't find a LoRA you know is there, add your Civitai API key in the settings - the site keeps some models hidden from signed-out users, and a key lets the lookup see them. The key stays on your computer and never goes into a workflow.

Install

Part of the Pixaroma pack: ComfyUI Manager → search Pixaroma → install → restart, or:

cd ComfyUI/custom_nodes
git clone https://gitlab.com/pixaroma/comfyui-pixaroma.git

No extra dependencies; your LoRAs live in the standard models/loras/. The pack's usual gotcha applies - hard-refresh the browser tab (Ctrl+Shift+R / Cmd+Shift+R) after an update, or the cached node visuals will make you think the update broke it.

If you know rgthree's Power Lora Loader, this is the same niche from a different family - both stack LoRAs with per-entry toggles and read trigger words, and plenty of people run both. The Pixaroma one adds the read-the-file trigger picker plus the triggers string output wired straight into your prompt. Default strengths are often too strong at 1.0; 0.5–0.8 is the common sweet spot, and the per-row arrows make nudging that painless.