ComfyUI Model Preset Pilot

🧭 ComfyUI Model Preset Pilot

A ComfyUI custom node module for managing model presets and configurations. This module automatically detects the model, loads/saves presets (sampler, scheduler, steps, CFG, clip-skip, seed, resolution), and generates/caches preview images per model.

Installation

1. Clone this repository into your ComfyUI custom_nodes directory:

   cd ComfyUI/custom_nodes
   git clone https://github.com/yourusername/ComfyUI-Model_preset_Pilot.git
   

2. Install dependencies (if not already installed):

   pip install -r ComfyUI-Model_preset_Pilot/requirements.txt
   

3. Restart ComfyUI

Features

• Automatic Model Preset Management: Load, save, and update presets for each model
• Preset Parameters: sampler, scheduler, steps, CFG, clip-skip, seed, resolution
• Preview Generation: Create and cache standardized preview images for each model
• Manual Preview Loading: Easily load custom images as model previews
• File Browser Integration: Upload preview images directly from your computer
• Multiple Operation Modes: Load, Save, Update
• Flexible Input Options: Works with both direct conditioning or CLIP+text

Usage

Basic Workflow

1. Add the "🧭 Model Preset Pilot" node to your workflow
2. Connect a model (checkpoint loader) to the "model" input
3. Set the mode to "load" to retrieve saved presets for this model
4. Use the output parameters in your workflow nodes

Modes

• Load Mode: Reads (or creates default) preset for the model and returns parameters + preview
• Save Mode: Overwrites the preset with current input values
• Update Mode: Updates only the fields provided, leaves others intact

Preview Generation

Automatic Preview Generation

To generate a preview image automatically:

1. Connect VAE and either:
   • CLIP + positive/negative prompts, or
   • Positive/negative conditioning
2. Set "generate_preview" to True
3. Optionally set "overwrite_preview" to True to regenerate existing previews

Manual Preview Loading

To manually load an image as a model preview:

1. Connect an image to the "image" input
2. Set "load_image" to True
3. Run the node to save the image as the model's preview

File Browser Upload

To upload an image directly from your computer:

1. Connect a model or specify a checkpoint name
2. Click the "Browse for preview image..." button that appears in the node
3. Select an image file from your computer
4. The image will be automatically uploaded and set as the model's preview

The preview images (whether loaded manually or uploaded via file browser) will be:

• Copied to the model's preview directory
• Displayed whenever the model is loaded
• Backed up if they already exist (as .png.bak)

File Storage

• Presets are stored as JSON files in: ComfyUI/user/model_presets/<model_id>.json
• Preview images are stored as PNG files in: ComfyUI/user/model_presets/previews/<model_id>.png

Example Workflows

Basic Preset Loading

1. Load checkpoint → Model Preset Pilot (mode: load) → KSampler (connect all parameters)
2. Use the returned parameters to automatically configure your workflow

Saving Custom Presets

1. Load checkpoint → Connect to Model Preset Pilot
2. Set mode to "save"
3. Input your preferred parameters
4. Run to save the preset for future use

Development

To contribute to this project:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

Project Structure

• model_preset_pilot.py: Main node implementation
• __init__.py: Node registration
• user/model_presets/: Directory for storing presets and previews