Lora Loader

Every big node suite eventually ships its own version of the basics, and Lora Loader is WAS's take on the same job ComfyUI's built-in LoraLoader does: apply a LoRA on top of your model (and usually CLIP) at a given strength. The README describes WAS's own loader family generally as "a set of ComfyUI loaders that also output a string that contains the name of the model being loaded" - the standing feature that separates WAS's loaders from the native ones is that extra string output carrying the filename, which is genuinely useful if you want to tag a saved image or filename with which LoRA actually produced it, wired through WAS's own token/text system elsewhere in the pack.

If you're only ever using one loader style in a graph, the native LoraLoader and this one land in roughly the same place functionally. The reason to reach for WAS's version specifically is that string output - skip it if you don't need to track which LoRA name produced a given result.

What actually matters when a LoRA "doesn't work"

This is general LoRA advice, not specific to this node, but it's the checklist worth running before assuming a loader is broken:

• Base model mismatch. A LoRA trained for one architecture doesn't transfer to another - a Flux LoRA on an SDXL checkpoint, or vice versa, simply won't apply correctly no matter which loader you use.
• Missing trigger word. Plenty of LoRAs need a specific word in the prompt to actually activate what they were trained for - check the model's page for it.
• Strength too high. The default full strength (1.0) is often more than you want; 0.5–0.8 is a common range to start walking down from if results look overcooked.

How to install it

ComfyUI Manager: search "WAS Node Suite," install, restart. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/WASasquatch/was-node-suite-comfyui

then, from inside that folder, install requirements - path/to/ComfyUI/python_embeded/python.exe -s -m pip install -r requirements.txt on portable, pip install -r requirements.txt on system Python. Restart ComfyUI; it's under WAS Suite. No model download of its own - it loads whatever LoRA file you already have in ComfyUI/models/loras/, same as the native loader.

Common issues & troubleshooting

If the LoRA doesn't seem to be having any visible effect, work through the general checklist above before suspecting the node itself - base-model compatibility and a missing trigger word are the two most common reasons a LoRA appears to "do nothing" regardless of which loader applied it. A useful diagnostic from the community: fix your seed, generate once with the LoRA and once without, otherwise identical settings, and compare directly rather than judging from memory across different seeds.

Beyond LoRA-specific issues, this node inherits the pack's general risk profile: WAS Node Suite has had no active development since the author marked it retired in December 2023, and the recurring complaint since then is the entire suite failing to import after a ComfyUI update - usually a shared-dependency version clash rather than a problem with this node specifically. If the whole suite goes "Import Failed," reinstall the requirements using your embedded Python interpreter (path/to/ComfyUI/python_embeded/python.exe -s -m pip install -r requirements.txt) rather than a system one - that's the fix that's actually worked for people stuck on it.