n8n-nodes-deepseek-chat-model

Developed and maintained by Jay Nguyen (Nguyễn Thiệu Toàn).

🛡️ Verified n8n Creator | 💼 CEO/Founder of GenStaff

Connect with me:

An optimized, community-verified n8n Chat Model node for DeepSeek (deepseek-chat, deepseek-reasoner, etc.) that solves the reasoning content tool-calling bug in n8n AI Agents and LangChain workflows.

Features

• Corrected Reasoning Logic: Automatically preserves and passes back the reasoning_content field in multi-turn tool-calling loops, preventing 400 Bad Request API errors.
• Thinking Mode Control: Easily toggle DeepSeek's thinking/reasoning process (Enabled/Disabled) and configure its thinking effort (high / max).
• AI Agent & LangChain Integration: Fully compatible with n8n AI Agent and Advanced AI nodes.
• Fully Custom model names: Support for standard models (deepseek-chat, deepseek-reasoner) as well as custom/self-hosted deployment models.

Installation

Go to Settings > Community Nodes in your n8n instance and install:

n8n-nodes-deepseek-chat-model

Credentials Configuration

1. Get your API Key from the DeepSeek Console.
2. In n8n, set up a new DeepSeek API credential:
   • API Key: Enter your DeepSeek API key.
   • Base URL Override: Default is https://api.deepseek.com, but can be replaced if using a proxy or local inference model (e.g. Ollama).

License

MIT