Self-hosted AI Starter Kit
Docker Compose template: n8n, Ollama, Qdrant and Postgres for local AI flows
What it is
An official n8n template that brings up a local low-code AI environment in one Compose project: n8n for workflows, Ollama for local models, Qdrant as vector store and PostgreSQL for data. It includes GPU and CPU profiles and a demo workflow. Good for anyone who wants to prototype private AI automations without cloud services.
Good for
- Prototype a local RAG workflow in n8n
- Summarise PDFs without sending data out
- Try AI agent nodes against a local model
What you need
Docker and Docker Compose; a GPU helps but is optional
Sources
- https://github.com/n8n-io/self-hosted-ai-starter-kit (project README, read 11 Nov 2026)
- https://api.github.com/repos/n8n-io/self-hosted-ai-starter-kit (GitHub API, read 13 Nov 2026)
The summary is ours, written from the README. Spotted an error? See how we work.