Extensions/ComfyUI_Dados_Nodes
ComfyUI Extension

ComfyUI_Dados_Nodes

ComfyUI_Dados_Nodes is a collection of custom nodes for ComfyUI, designed to enhance functionality and provide integration with various services, including Pinterest. This privacy policy explains how these nodes handle user data. NOTE: a/privacy_policy

By DadoIrie·Created 2 years ago·Updated 8 days ago· 10
dadoirie/ComfyUI_Dados_Nodes
Nodes15
On cloudLocal install
CategoryDado's Nodes/replicate/Image, Dado's Nodes/Image
Stars10
Updated8 days ago

Nodes (15)

Replicate (0xdino/cyberrealistic-pony-v125)

The Pony-photoreal model you rent instead of run

Dado's Nodes/replicate/Image
Base64 to Image

The plumbing node that turns base64 back into pixels

Dado's Nodes/Image
CSV Multi DropDown

One CSV, a row of dropdowns, zero text-node spaghetti

Dado's Nodes/Text & Prompt
Image Batcher

Stacking images into one batch, bars and all

Dado's Nodes/Image
Image to Base64

Turn any ComfyUI image into a string the rest of the world can read

Dado's Nodes/Image
Memory Storage

A scratchpad for strings that survives across runs

Dado's Nodes/Memory Storage
Multiline String

The 20-line node that every prompt workflow ends up needing

Dado's Nodes/Text & Prompt
Preview Image (Dados Nodes)

One preview node that handles both images and video paths

Dado's Nodes/Image
Pinterest Image Node

The Pinterest scraper node that feeds reference boards into your graph

Dado's Nodes
Python String Function

Run a tiny sandboxed Python function inside your workflow

Dado's Nodes/Utils
Video Last Frame

Grab the last frame of a video and hand it to your graph

Dado's Nodes/Video
Video Merge

Concatenate two mp4s without the duplicated frame stutter

Dado's Nodes/Video
Wildcards Processor

Actually expand your wildcards, instead of pasting them into the sampler

Dado's Nodes/Text & Prompt
Replicate (qwen/qwen-image-edit-plus)

Tell Qwen what to change, skip the mask entirely

Dado's Nodes/replicate/Image
Replicate (wan-video/wan-2.2-i2v-fast)

Wan 2.2 i2v as a service — even the speed-LoRA debate is built in

Dado's Nodes/replicate/Video
Readme

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 selection and duplicate filtering
  • CSV Multi DropDown - Parse CSV rows into dynamic dropdowns with duplicate filtering and random selection
  • 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

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.