I have updated my N8N Self-Hosted Docker compose library to include SearXNG [http://docs.searxng.org/]. SearXNG is an internet metasearch engine that aggregates results from up to 246 search services. It is privacy focused and does not track or share data.
This is an excellent addition to N8N when coupled with Local AI. You get enhanced results by utilizing web search data with your AI workflow
Here is an example Chat AI workflow with an SearXNG, It allows the agent to query data from the web to answer questions.
When used SearXNG as part of a workflow, the AI agent can perform searches and look up information during the AI session. In this example, I instructed a simple AI agent to “tell me about the blog jimiz.net.”
The AI agent was running Ollama locally and did not have any information in the LLM (Large Language Model) (Qwen2.5:latest) about the website jimiz.net. Instead, it utilized SearXNG to search for and retrieve information. In the image, I share the logs showing that it queried about jimiz.net and then returned the answer.
As you experiment and build more AI agents, you will begin to utilize and chain tools together. SearXNG is a powerful tool to incorporate into your solutions.
In my next blog or video, I will share some of the MCP servers I utilize.
Here is my self-hosted N8N Solution
https://github.com/jimiz12/n8n-docker-postgresql?utm_source=chatgpt.com