n8n + ApexApi

How to Crawl a Website into an AI Knowledge Base with n8n

This workflow crawls a whole site into markdown, then an AI model explains what it is — the core pattern for building a chatbot or RAG system from a website. Cap the page count to control cost; feed the pages into a vector store to go all the way to a knowledge base.

What you'll build

  • Crawl a full site to clean markdown as one async job
  • A page limit that caps your cost per crawl
  • AI output describing the site — or feed pages into a vector DB for RAG
  • Per-page pricing; failed pages are free

Set it up in 4 steps

1

Get a free ApexApi key

Create a key at apexapi.dev — no card required. It starts with ak-.
2

Install the community node

In n8n: Settings → Community Nodes → Install n8n-nodes-apexapi (0.2.0+).
3

Import this template

Download the workflow, then in n8n use Workflows → Import from File. It uses the ApexApi Context node's Read a Site operation.
4

Add your key and run

Open any ApexApi node, create an ApexApi account credential, paste your key, then click Test workflow.

FAQ

How big a site can I crawl?

Up to 500 pages per job. Set the Max Pages field to whatever you need — you only pay for pages returned.

Can I build a real chatbot from this?

Yes — the crawl returns a pages[] array; pipe it into an n8n vector-store node and connect an AI Agent for full RAG.

What does it cost?

$0.005 per page crawled. Failed or empty pages are free, and the job is capped at your page limit.

More n8n templates

See all node operations in the n8n integration docs.