This repository provides a comprehensive suite of 163 custom nodes for ComfyUI, aimed at enhancing text, image, and video manipulation, as well as workflow management and integration with external AI tools like Ollama and Text-to-Speech. It offers advanced features for randomization, conditional logic, looping operations, and detailed control over generation processes, empowering users to build complex and dynamic AI workflows.

How It Works

The nodes are designed to extend ComfyUI's capabilities by offering granular control over various aspects of AI generation. Key features include sophisticated text manipulation with advanced syntax for randomization and variable management, extensive looping mechanisms for iterating through data, models, or parameters, and integrated support for external services like Ollama for conversational AI and TTS for speech synthesis. The project also includes nodes for direct downloading from Civitai and Hugging Face, simplifying model and LoRA management.

Quick Start & Requirements

• Installation: Typically involves placing the repository folder into ComfyUI's custom_nodes directory.
• Dependencies: Python environment with libraries like ollama, pydub, opencv-python, ffmpeg-python, civitai-py, fal_client, sounddevice, langdetect, spacy, textblob, dill. For specific nodes, faster-whisper may require manual installation.
• Setup: Requires ComfyUI to be installed and running. Detailed installation instructions for Windows and Linux are provided, including recommendations for using Python virtual environments.
• Resources: Usage of certain nodes, particularly those involving TTS or large model operations, may require significant VRAM and processing power.

Highlighted Details

• Over 163 distinct custom nodes categorized for easy access.
• Advanced text generation capabilities with custom syntax for randomization and conditional logic.
• Extensive looping and selection nodes for models, LoRAs, text, and images.
• Direct integration with Ollama for chat and vision tasks.
• Support for Text-to-Speech (TTS) with local voice generation.
• Nodes for downloading models and LoRAs directly from Civitai and Hugging Face.
• Video manipulation nodes including splitting, merging, and synchronization.
• Workflow control nodes for pausing, resuming, and managing execution flow.

Maintenance & Community

The project is under very active development with frequent updates and new features. The author encourages issue reporting and community feedback. Support is available via Ko-fi.

Licensing & Compatibility

The README does not explicitly state a license. Compatibility with commercial use or closed-source linking would depend on the specific, unstated license.

Limitations & Caveats

The project is explicitly marked as "Very active development. Work in progress." Some nodes are experimental or have known limitations (e.g., FFMPEG nodes may still require improvements, TTS node not tested on Windows, faster-whisper compatibility issues with Python 3.13). Users should anticipate potential breaking changes and bugs.