Eden.art Custom Node Suite for ComfyUI

A collection of powerful custom nodes and workflows developed by Eden

This repository contains a comprehensive suite of specialized ComfyUI nodes designed to enhance our generative workflows. While some nodes may not be fully documented here yet, they are actively used in [our workflows repository](https://github.com/edenartlab/workflows) which powers many of the creative AI tools on **https://www.eden.art/** ## 🤝 Contributing We welcome contributions! If you want to contribute custom ComfyUI workflows as tools into our creative AI agent ecosystem, please checkout our in-production workflows repository: https://github.com/edenartlab/workflows ## 🌟 Featured Nodes ### GPT4 Node

**Call GPT4 for Text Completion** - Versatile node that provides seamless integration with the OpenAI API - Configuration: Place a `.env` file containing your API key in your root ComfyUI folder --- ### GPT4 Vision Node

**Image Captioning and Visual Understanding** - Interprets images and generates detailed, contextual descriptions - Configuration: Requires an `.env` file with your OpenAI API key in the root ComfyUI folder --- ### Load Random Images LoadRandomImage Node

**Dynamic Random Image Selection for Automated Workflows** - Queue your prompt multiple times to process different inputs through the same workflow - Automatically processes multiple images to maintain consistent aspect ratio and resolution --- ### Color Clustering Mask Generator maskfromrgb_kmeans Node

**Generate Precise Masks Using Color Clustering** - Implements KMeans clustering algorithm to produce accurate segmentation masks - Particularly effective for creating AnimateDiff masks directly from source videos --- ### DepthSlicer DepthSlicer Node

**Create Targeted Masks from Depth Maps** - Creates segmentation along the z-axis to produce specific "depth slices" - Ideal for creative animations and precise inpainting applications --- ### 3D Parallax Zoom Parallax Zoom Node

**Create Immersive Three-Dimensional Effects** - Uses depth map and image data to generate Deforum-style 3D-zoom parallax videos - Adds spatial dimension and dynamic movement to static images --- ### CLIP Interrogator CLIP Interrogator Node

**Extract Detailed Text Descriptions from Images** - Based on the [clip-interrogator](https://github.com/pharmapsychotic/clip-interrogator) project - Key features: - Optional model persistence in GPU memory with `keep_model_alive` parameter - Configurable BLIP caption inclusion via `prepend_BLIP_caption` parameter - Automatic prompt saving with `save_prompt_to_txt_file` option - Installation note: If auto-download fails, manually clone https://huggingface.co/Salesforce/blip-image-captioning-large into ComfyUI/models/blip --- ### VAEDecode to Folder VAE Decode to Folder Node

**Direct-to-Disk VAE Decoding** - Saves decoded images directly to a specified directory - Enables rendering of extended video sequences with AnimateDiff - Note: Requires manual video compilation using ffmpeg afterward --- ### Random Number Sampler Sample Uniform Random numbers

**Sampling** - Visualizes the sampled number on top of the node using custom javascript --- ### Enhanced SaveImage Save Image Node

**Advanced Image Preservation Options** - Add timestamps to filenames for better organization - Automatically save the complete pipeline as a JSON file - Easily reference prompts and settings without loading the entire pipeline --- > **Note:** Some nodes in this collection are still under active development and may not be fully functional or documented. ## 🔧 Installation ```bash cd ComfyUI/custom_nodes/ git clone https://github.com/edenartlab/eden_comfy_pipelines.git cd eden_comfy_pipelines pip install -r requirements.txt ``` ## 📜 License This project is licensed under the MIT License - see the LICENSE file for details.

ComfyUI Node: Load_Embeddings_From_Folder

Category

Inputs

Outputs

Extension: Eden.art nodesuite

Run ComfyUI workflows in the Cloud!