DeepSeek AI Integration Quick Start: Chat & Reasoning Agent

Unlock Advanced AI Capabilities with this DeepSeek Quick Start Agent

This n8n workflow is your launchpad for integrating DeepSeek's cutting-edge language models into your automations. It's designed as a comprehensive quick start guide, showcasing multiple methods to connect and interact with deepseek-chat (V3 for general conversation) and deepseek-reasoner (R1 for complex reasoning tasks). Whether you prefer cloud APIs, LangChain's structured approach, or local models via Ollama, this agent has you covered.

Use this AI Agent to:

• Rapidly prototype conversational AI applications.
• Explore the reasoning power of DeepSeek-R1 for data analysis or problem-solving.
• Learn best practices for calling LLMs within n8n.

Key Features & Benefits

• Multiple DeepSeek Models: Demonstrates usage of both deepseek-chat (V3) and deepseek-reasoner (R1).
• Versatile Connection Methods:
  • Direct API Calls: Examples using n8n's HTTP Request node for deepseek-chat and deepseek-reasoner.
  • LangChain Integration: Utilizes the LMChatOpenAI node, leveraging DeepSeek's OpenAI API compatibility.
  • Local Model Support: Shows how to connect to DeepSeek models running locally via Ollama using the LMChatOllama node.
• Conversational Agent Setup: Includes a LangChain agent with window buffer memory for building interactive chat experiences.
• Practical Examples: Pre-configured nodes and sticky notes explain each approach, making it easy to adapt and extend.
• Empowers AI-Driven Automation: Quickly add sophisticated natural language understanding, generation, and reasoning abilities to any n8n workflow.

• An n8n instance (Cloud or self-hosted).
• DeepSeek API Key (obtainable from platform.deepseek.com/api_keys).
• (Optional) Ollama setup with DeepSeek models (e.g., deepseek-r1) downloaded for local execution.
• n8n credentials configured for DeepSeek API (e.g., 'Generic Credential Type: HTTP Header Auth' for HTTP nodes, or an OpenAI compatible credential for LangChain nodes pointing to DeepSeek's API endpoint).
• (Optional) n8n credentials for your local Ollama instance if using Ollama nodes.

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Configure DeepSeek API Access:
   • For the 'DeepSeek' LangChain node (using OpenAI compatibility): Create or select an n8n credential for OpenAI API. In its configuration, set the 'Base URL' to https://api.deepseek.com/v1 and use your DeepSeek API key.
   • For the 'DeepSeek JSON Body' / 'DeepSeek Raw Body' HTTP Request nodes: Create or select an n8n 'HTTP Header Auth' credential. Add a header with 'Name' as Authorization and 'Value' as Bearer YOUR_DEEPSEEK_API_KEY (replace YOUR_DEEPSEEK_API_KEY with your actual key).
4. (Optional) Configure Ollama:
   • Ensure Ollama is running on your system or a reachable server.
   • Pull the desired DeepSeek model (e.g., ollama pull deepseek-r1:14b).
   • In the 'Ollama DeepSeek' node, select or create an n8n credential for your Ollama instance, specifying its base URL (e.g., http://localhost:11434).
5. Review the different connection examples provided (LangChain Agent with memory, Basic LLM Chain with Ollama, direct HTTP Requests). The workflow is structured with sticky notes explaining each section.
6. The 'When chat message received' trigger is pinned with example data. You can connect it to any of the demonstrated DeepSeek interaction paths to test.
7. Customize the prompts, models (deepseek-chat for chat, deepseek-reasoner for reasoning), and parameters within the nodes as needed for your specific use case.
8. Activate the workflow to use your DeepSeek AI Agent.