What it does
A self-hosted n8n workflow that runs a conversational agent through the LangChain Code node, with Google Gemini as the language model. You control the prompt and the length of the conversation memory. It avoids the reserved tool-calling functionality of n8n's built-in agent, which can add token use. The LangChain Code node only works on self-hosted n8n.
Use cases
- 01Building an internal chat assistant with a fixed prompt
- 02Testing a conversation in the n8n editor
- 03Swapping the Gemini model for another provider
Connects
LangChain Code · Simple Memory · Google Gemini Chat Model