Monitor AI chat interactions with Gemini 2.5 and Langfuse tracing

This workflow contains community nodes that are only compatible with the self-hosted version of n8n.

How it works

This workflow is a simple AI Agent that connects to Langfuse so send tracing data to help monitor LLM interactions.

The main idea is to create a custom LLM model that allows the configuration of callbacks, which are used by langchain to connect applications such Langfuse.

This is achieves by using the "langchain code" node: - Connects a LLM model sub-node to obtain the model variables (model name, temp and provider) - Creates a generic langchain initChatModel with the model parameters. - Return the LLM to be used by the AI Agent node.

📋 Prerequisites - Langfuse instance (cloud or self-hosted) with API credentials - LLM API key (Gemini, OpenAI, Anthropic, etc.) - n8n >= 1.98.0 (required for LangChain code node support in AI Agent)

⚙️ Setup

1. Add these to your n8n instance: bash Langfuse configuration LANGFUSESECRETKEY=yoursecretkey LANGFUSEPUBLICKEY=yourpublickey LANGFUSEBASEURL=https://cloud.langfuse.com or your self-hosted URL