Skip to content

Latest commit

 

History

History
128 lines (96 loc) · 3.38 KB

README.md

File metadata and controls

128 lines (96 loc) · 3.38 KB

ComfyUI DeepSeek_R1 Chat Node

A custom node for ComfyUI that integrates DeepSeek's powerful chat and instruction API, enabling seamless AI interactions within your ComfyUI workflows.

image

Features

  • Dual Operating Modes

    • Chat mode with conversation memory
    • Single-turn instruction mode
  • Vision Integration

    • Support for image context through vision descriptions
    • Seamless integration with ComfyUI's visual workflows
  • Advanced Configuration

    • Customizable max token limit (up to 8192 tokens)
    • Optional reasoning display
    • Configurable chat history management
    • Built-in rate limiting protection

Installation

  1. Clone this repository into your ComfyUI custom nodes directory:
cd ComfyUI/custom_nodes
git clone https://github.com/ShmuelRonen/comfyui-deepseek-chat.git
  1. Get your DeepSeek API key:

  2. Edit theconfig.json file in the node directory:

{
    "deepseek_api_key": "your-api-key-here"
}

Important

Check API availability on this site: [DeepSeek API status](https://platform.deepseek.com/

Usage

Node Configuration

The node provides several input parameters:

  • mode: Choose between "chat" and "instruct"
  • prompt: Your input text/question
  • max_tokens: Maximum response length (1-8192)
  • show_reasoning: Toggle reasoning display
  • clear_history: Option to reset conversation
  • remember_chat: Enable/disable chat history persistence
  • vision_description: Optional image context description. This can be used to:
    • Provide image descriptions for visual context
    • Connect with image generation nodes
    • Enable vision-language tasks
    • Process image analysis results

Example Workflow

  1. Add the "DeepSeek-V3 Chat/Instruct" node to your workflow
  2. Configure the node parameters
  3. Connect it to other nodes as needed
  4. Execute the workflow

Outputs

The node provides two outputs:

  • latest_response: The most recent AI response
  • full_chat_history: Complete conversation history

Rate Limiting

The node implements a 1-second minimum interval between requests to comply with API limits.

Error Handling

  • Comprehensive error logging
  • Graceful handling of API failures
  • Clear error messages in node outputs

Technical Details

API Integration

API Endpoint: https://api.deepseek.com/v1/chat/completions
Model: deepseek-reasoner
  • Supports full chat context with message history
  • Handles authentication via API key
  • Implements rate limiting protection

File Structure

comfyui-deepseek-chat/
├── deepseek_chat_node.py
├── config.json
├── chat_history.json
└── README.md

Dependencies

  • Python 3.6+
  • requests library
  • ComfyUI environment

Contributing

  1. Fork the repository
  2. Create your feature branch
  3. Commit your changes
  4. Push to the branch
  5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

  • Built for ComfyUI
  • Powered by DeepSeek's API
  • Inspired by community needs for AI integration in visual workflows