This repository provides a custom node plugin for ComfyUI that facilitates the configuration and storage of parameters related to large language models (LLMs) such as OpenAI, Kimi, and DeepSeek. Users can input essential details like API Key, Base, Version, and Model, which are then automatically saved locally and set as environment variables for convenient future access.
- Enables users to select LLM types and input relevant parameters through a form interface.
- Automatically saves configuration data to a designated JSON file for persistence.
- Sets environment variables to streamline access for subsequent nodes or scripts.
Context
The ComfyUI-Custom-Node-Config tool is designed to enhance the ComfyUI framework by providing a structured way to manage LLM configurations. Its primary purpose is to simplify the process of entering and storing API-related information, making it easier for users to integrate various LLMs into their workflows.
Key Features & Benefits
This tool offers a user-friendly form input system that allows users to select their desired LLM and enter necessary parameters, such as API Key and Version. The automatic saving of configurations to a JSON file ensures that user settings are preserved between sessions, while the environment variable setup facilitates easy retrieval of these parameters by other nodes or scripts in ComfyUI.
Advanced Functionalities
The plugin includes a class called FormSubmitNode, which encapsulates the logic for handling user input, saving configurations, and managing environment variables. Users can extend the functionality by modifying the input types and related methods within the plugin's code, allowing for customized field additions as needed.
Practical Benefits
By integrating this tool into ComfyUI, users gain improved workflow efficiency, as they no longer need to manually input API parameters repeatedly. The automatic configuration saving and environment variable management enhance control and reduce the likelihood of errors, ultimately leading to a smoother experience when working with LLMs.
Credits/Acknowledgments
This project is maintained under the MIT License, and contributions to its development can be made through issues or pull requests on the repository.