LoRA Txt Loader — ComfyUI Custom Node

A ComfyUI custom node that loads LoRA trigger words from a local .txt file — fully offline, fully customizable, no metadata dependency.

Preview

LoRA Txt Loader — Tree Browser

The main node with trigger words auto-loaded from a .txt file. Click the LoRA name button to open the tree browser. Right-click to manage tag versions.

Tree-Style Folder Browser

Browse thousands of LoRAs in a collapsible tree panel. Only one folder per level expands at a time (accordion behavior). Built-in search filters across all files instantly.

LoRA Txt Loader (Dropdown)

The dropdown variant — same auto-fill behavior, classic ComfyUI interface.

How It's Different

Most existing LoRA loader nodes get trigger words from one of two sources:

• Model metadata baked into the .safetensors file (often incomplete or missing entirely)
• Civitai API — requires internet access, and only works for models publicly hosted on Civitai

LoRA Txt Loader takes a different approach.

You maintain a plain .txt file next to each LoRA, and the node reads from it directly. This means:

• Works completely offline
• Works for self-trained models that have never been uploaded anywhere
• Trigger words are exactly what you wrote — no guessing, no scraping
• You're in full control

Features

• 📁 Tree-style folder browser — browse your LoRA folders in a collapsible tree panel, with accordion behavior so only one folder opens at a time, keeping the view clean
• 🔍 Built-in search — type to filter across all LoRAs instantly, supports up to 10,000+ files without lag
• 📌 Remembers your position — reopening the panel automatically expands and scrolls to your currently selected LoRA
• 📄 Auto-loads trigger words from a .txt file with the same name as the LoRA
• ✏️ Editable text box — modify trigger words directly in the node before running
• 🔄 Auto-fills on LoRA switch — when you select a different LoRA, the text box updates automatically
• ↩️ Reset button — one click to restore the original .txt content if you've edited it
• ⟳ Refresh button — reload the LoRA list without restarting ComfyUI (useful after downloading new LoRAs)
• 📐 Remembers node size — manually resizing the node is preserved when switching LoRAs
• 🔌 Drop-in replacement for the native Load LoRA node — same inputs, same outputs, plus extras
• 🏷️ Tag version management — save, activate, and manage multiple versions of LoRA trigger words with custom names and notes

New: LoRA Tag Version Management

Right-click on a LoRA to save, manage, and switch between different versions of trigger words (tags).

How It Works

Right-click menu options:

• Save LoRA Tags Version — Save current prompt text as a new version
• LoRA Tags Version Manager — Open the version manager to view, activate, or delete versions

File Structure

Versions are stored in a [lora_name].versions/ folder alongside the LoRA model:

my_character.safetensors
my_character.txt              ← Currently active version (always same as LoRA name)
my_character.versions/
├── manifest.json            ← Version metadata
├── _original.txt            ← Default version (from original my_character.txt)
├── 2026-06-30 22-02-38_v1.txt
├── 2026-06-30 22-02-42_v2.txt
└── 2026-06-30 22-03-08_v3.txt

• _original.txt — the default version (what was in the original .txt file)
• YYYY-MM-DD HH-MM-SS_versionname.txt — custom versions with timestamp and name
• manifest.json — stores version names, notes, and metadata
• my_character.txt — always represents the currently active version

First Time Setup (No Versions Yet)

Scenario A: LoRA has a corresponding .txt file and it's not empty

• The .txt content is automatically saved as _original.txt in the .versions/ folder (the default version)
• You can now create additional custom versions

Scenario B: No .txt file or .txt is empty

• The current text in the text box is saved as _original.txt (the default version)
• Then you can add more versions

Managing Versions

Once you have at least one version saved:

1. Right-click on the LoRA → "Save LoRA Tags Version" to create a new version

   • Enter a custom version name
   • System creates: YYYY-MM-DD HH-MM-SS_versionname.txt in the .versions/ folder
   • Add optional notes/description (stored in manifest.json)

2. Version Manager → Access all saved versions:

   • Activate a version to make it active (the version's content is copied to my_character.txt)
   • Delete a version (cannot delete the currently active or default _original.txt)

3. Switching versions:

   • Activating a version replaces the same-named .txt file (e.g., my_character.txt) with that version's content
   • After switching versions, click Refresh LoRA to update the downstream LoRA Txt Loader (Dropdown) node if you're using it
   • The node always reads from the same-named .txt file

Example Workflow

LoRA: my_character.safetensors
├── Default Version (_original.txt)
│   └── "white hair, blue eyes, school uniform"
├── Version: Summer Outfit (2026-06-30 22-02-38_Summer_Outfit.txt)
│   └── "white hair, blue eyes, summer dress"
└── Version: Winter Outfit (2026-06-30 22-02-42_Winter_Outfit.txt)
    └── "white hair, blue eyes, winter coat, scarf"

Switch between versions as needed — each saves your custom trigger words separately.

Nodes

1. LoRA Txt Loader (recommended)

The main node. Uses a tree-style folder browser with built-in search to select a LoRA. Click a folder to expand/collapse it — only one folder per level stays open at a time. Auto-fills the prompt text box from the adjacent .txt file. Includes a Reset button and a Refresh button to pick up newly downloaded LoRAs without restarting ComfyUI. Right-click to manage trigger word versions.

2. LoRA Txt Loader (Dropdown)

Same as above but uses the standard ComfyUI dropdown instead of the tree browser. Useful if you prefer the classic interface.

3. LoRA Txt Loader (From Path)

Takes a direct file path string as input instead of a dropdown. Useful for dynamic workflows where the LoRA path is generated by another node. Reads trigger words from the adjacent .txt file automatically.

| Input | Description | |-------|-------------| | model | Connect from checkpoint loader | | clip | Connect from checkpoint loader | | lora_path | Full path to the .safetensors file | | strength_model | LoRA model strength | | strength_clip | LoRA CLIP strength |

4. Txt File Loader

Loads a standalone .txt file from the ComfyUI/input/texts/ folder into an editable text box. Useful for storing reusable prompt snippets, style templates, or shared trigger word sets separately from any specific LoRA.

Place your .txt files in:

ComfyUI/input/texts/
├── my_style.txt
├── base_prompt.txt
└── ...

Installation

cd ComfyUI/custom_nodes
git clone https://github.com/kwokkakiu233/comfyui-lora-txt-loader

Restart ComfyUI after installation.

Usage

1. Prepare your .txt files

Place a .txt file with the same base name as your LoRA in the same folder:

ComfyUI/models/loras/
├── my_character.safetensors
├── my_character.txt          ← trigger words go here
├── subfolder/
│   ├── my_pose.safetensors
│   └── my_pose.txt

The .txt file can contain anything — single tags, comma-separated words, full prompt snippets, whatever works for your workflow.

2. Add the node

Find it under loaders → LoRA Txt Loader

3. Connect and use

| Input | Description | |-------|-------------| | model | Connect from checkpoint loader | | clip | Connect from checkpoint loader | | lora_name | Click to open the tree browser and select a LoRA | | strength_model | LoRA model strength | | strength_clip | LoRA CLIP strength | | positive_prompt | Auto-filled from .txt; edit freely |

| Output | Description | |--------|-------------| | model | Patched model | | clip | Patched CLIP | | positive_prompt | Your (edited) trigger words — connect to downstream text nodes | | lora_path | Full path to the loaded LoRA file |

4. Using positive_prompt in your workflow

The positive_prompt output is a plain string — you can wire it into your workflow in several ways:

Option A — Direct encode (simplest) Connect positive_prompt directly to a CLIP Text Encode node. The trigger words become your entire positive prompt.

Option B — Append to a base prompt (recommended) Use a Text Concatenate node to merge trigger words with your own prompt:

[Your base prompt] + [positive_prompt from LoRA Txt Loader]
        ↓
  Text Concatenate
        ↓
  CLIP Text Encode
        ↓
    KSampler (positive)

This lets you write a scene description once and swap LoRAs freely without rewriting your prompt each time.

Option C — Convert to conditioning Feed the output into a Text to Conditioning node if your workflow uses a conditioning-based pipeline.

5. Refresh without restarting

After downloading a new LoRA, click ⟳ Refresh LoRA List on the node. The tree browser will immediately include the new file — no need to restart ComfyUI.

6. Managing Tag Versions

Right-click on any LoRA to:

• Save LoRA Tags Version — create a new version of trigger words with a custom name and optional notes
• LoRA Tags Version Manager — view, activate, or delete existing versions

See LoRA Tag Version Management section above for details.

Why a .txt file?

If you train your own LoRAs, you already know what triggers them — the metadata inside the file is often empty or auto-generated garbage. A plain text file is the simplest possible way to keep notes next to a model. No tools required, editable in Notepad, survives any model conversion or re-export.

Bug Fixes

v1.x

• Fixed: Txt File Loader node now correctly locates .txt files in some edge cases where the file lookup previously failed

Requirements

• ComfyUI (Desktop or server version)
• Python 3.x (included with ComfyUI)
• No additional dependencies

License

MIT