Can I run n8n on a NAS?
Yes. n8n connects apps and services into automated workflows, such as watching a folder, summarising new documents with a local model and emailing the summary. Its AI nodes can call Ollama or another OpenAI-compatible endpoint on the same NAS, so the whole workflow stays private.
Which systems support it?
ZimaOS lists n8n in its app store, which is the only vendor listing we found. QNAP says n8n is being integrated on the QAI-h1290FX. Everywhere else it runs in Docker without a vendor guide, so most rows below show Unverified support.
What RAM is needed?
n8n's documentation does not state a minimum. It is light on its own; memory planning depends on the models your workflows call. On a NAS that also runs a local model, size the memory for the model first.
Which architecture works?
The official image, docker.n8n.io/n8nio/n8n, is published for amd64 and arm64, so it runs on ARM NAS systems too, as long as the workflows do not need a local model.
How do I install it?
Create a container from the official image in your vendor's container tool, with a volume for n8n's data so workflows survive updates, and open its web interface. Add an Ollama credential pointing at the Ollama container to use local models in AI nodes.
What can you automate with local AI?
Common examples are summarising new files in a shared folder, sorting incoming email by topic, extracting fields from invoices or forms, drafting replies for review, and posting a daily digest of new documents to a chat channel. Each step that needs language understanding calls the local model; the rest is ordinary automation. Keep workflows that touch personal data on the NAS rather than routing them through cloud services.
Which systems should I buy for n8n with local AI?
The workflows are light; the models are not. The picks below combine confirmed Docker support with strong local AI scores and live UK prices.
Common questions
Is n8n free to self-host?
n8n is available to self-host under its fair-code licence; check the licence terms for business use.
Can n8n use local models?
Yes, through its Ollama and OpenAI-compatible nodes pointed at a model on the NAS.
