Load Lora's Trigger Word Automatically into Prompt Text

[RyuuRyu]

Hello, I'm new in ComfyUI. Just moved from A111 a week ago. Everything runs perfectly but then I notice that the effect that the Lora affect the image was lesser in ComfyUI than in A111. I thought I connect the Nodes incorrectly but then I remembered that some Lora needs trigger words. And I've been just load the Lora using Lora Loader all this time but didn't consider about the trigger words.

So, I search a way to load the Lora's trigger word into the CLIPTextEncoder, I found that the Custom Nodes called Lora Tags Loader allow me to load Lora into text instead. But it only Loads the Lora and doesn't automatically include the trigger word. I also found out that I could actually see information of Lora in CivitAI by using the Power Lora Loader by Rgthree. It does give me some information including the trigger words. But then I was confused about where to add the trigger word.

So, my question is,

Is there a way to not just Load Lora but also with its Trigger Words? Or do I even still need to add trigger words in ComfyUI since Lora was load with its Loader?

[SamSeenX]

Hi @RyuuRyu! Welcome to the ComfyUI nodes! 🧶

Short answer: Yes, you usually still need to add the trigger word to your text prompt.

Here is why:

• The LoraLoader node patches the model weights (it teaches the model how to draw the thing).
• The Trigger Word in your CLIPTextEncode prompt acts as the switch to tell the model "Draw that thing I just taught you."

Without the trigger word, the effect is often much weaker or invisible, which is exactly unmatched with what you experienced coming from A1111 (where some extensions might have auto-inserted it).

For automation:
If you want it fully automatic, check out the Civitai Helper custom nodes. They can fetch metadata and sometimes have features to auto-append trigger words, but honestly, simply typing the trigger word into your Positive Prompt is the most reliable way to ensure you get the exact effect you want!