2.Self-Hosted AI: n8n + Ollama, local AI workflows on your Mac
If you want AI agents running on your own machine, with your own models, and no data leaving your computer, this is the article :). This is part three of the series. In part one we set up PostgreSQL, and in part two we covered the LLM concepts (models, parameter, quantization, context, capabilities, VRAM). Today we put them to work: n8n for the workflows and Ollama for the models. One…
This story provides a step-by-step guide to setting up a local AI workflow using n8n and Ollama on a Mac machine. The process begins with installing Docker and then moving on to setting up n8n using the Self-hosted AI Starter Kit, which includes PostgreSQL, Ollama, Qdrant, and n8n. The prerequisites, such as installing Docker and a model like gemma4, are explained before moving on to connecting Ollama to n8n.
The guide explains how to create credentials in n8n for Ollama, set the base URL, and use it within a Chat Trigger node. The story also covers the use of tools within the AI Agent node and mentions that cloud API keys are only necessary for cloud models, not for local ones. Finally, it hints at building a real-world application, a document assistant using the Qdrant vector store, which will be covered in a subsequent article.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written; read the original for the full account.
