ComfyUI Dados Nodes

A collection of custom nodes for ComfyUI featuring AI vision models, advanced text processing tools, and wildcard prompt utilities.

Features

🤖 AI Vision Models

• SmolVLM Node - Lightweight multimodal model for image description and analysis
• JoyTagger - Advanced image tagging with customizable tag counts
• MiaoshouAI Tagger - Florence-2 based prompt generation and captioning

📝 Text Processing

• Dynamic Text Concatenate - Flexible text combination with custom delimiters
• Text DropDown - Interactive dropdown selections with random options
• Multiline String - Simple multiline text input handling

🎲 Wildcard System

• Wildcard Prompt Editor - Advanced wildcard editing with nested selections
• Wildcards Processor - Process wildcards with seed control and attention support

Note: Use Wildcard Prompt Editor output with Wildcards Processor for random wildcard processing.

Installation

Option 1: ComfyUI Manager (Recommended)

Simply search for "Dados Nodes" in ComfyUI Manager and install.

Option 2: Manual Installation

1. Navigate to your ComfyUI custom nodes directory:

   cd ComfyUI/custom_nodes/
   

2. Clone this repository:

   git clone https://github.com/dadoirie/ComfyUI_Dados_Nodes.git
   

3. Install dependencies:

   cd ComfyUI_Dados_Nodes
   pip install -r requirements.txt
   

4. Restart ComfyUI

Node Details

SmolVLM Image Describer

• Supports multiple model sizes (256M, 500M)
• Both SmolVLM and SmolVLM2 variants available
• Automatic model downloading from HuggingFace
• Customizable prompts and token limits

JoyTagger

• High-quality image tagging
• Adjustable tag count (1-100)
• Automatic model downloading
• Clean tag processing and formatting

MiaoshouAI Tagger

• Multiple instruction types: GENERATE_TAGS, CAPTION, DETAILED_CAPTION
• Florence-2 base and large models
• Version 1.5 and 2.0 support
• Configurable token limits

Wildcards Processor

• Seed-based randomization
• Attention generator support
• Compatible with dynamic prompts syntax
• Handles random wildcard selection from Wildcard Prompt Editor output or any other text/string source

Wildcard Prompt Editor

• Interactive web-based editor
• Nested wildcard support
• Real-time preview
• Selection persistence
• Connect output to Wildcards Processor for random wildcard processing

Requirements

• dynamicprompts - For wildcard processing functionality

Model Storage

Models are automatically downloaded to:

ComfyUI_Dados_Nodes/models/
├── smolvlm-256m-instruct/
├── smolvlm-500m-instruct/
├── smolvlm2-256m-video-instruct/
├── smolvlm2-500m-video-instruct/
├── florence-2-base-promptgen-v1.5/
├── florence-2-large-promptgen-v1.5/
├── florence-2-base-promptgen-v2.0/
├── florence-2-large-promptgen-v2.0/
└── joytag/

Known Issues

• SmolVLM2 Models: SmolVLM2-256M-Video-Instruct and SmolVLM2-500M-Video-Instruct are currently not working. This will be fixed in upcoming releases.

License

MIT License - see LICENSE file for details.

Contributing

Issues and pull requests are welcome on GitHub.

---

Note: This project is actively developed. Some features may change between versions.