Customer support chatbot on a $15-25/month VPS (Ollama + n8n)
A small-business guide to a self-hosted support bot: Ollama with Llama 3.2 3B answers from an FAQ knowledge base inside an n8n workflow, escalates to a human when unsure, and works from a web widget or support inbox.
Solo2-10 peopleCustomer supportUnder $1kSensitivity: Standard
Hardware
| Item | Approx. price (as of 2026-10-16) |
|---|---|
| On sale −6% MSI GeForce RTX 5060 Ti 16GB Ventus 3X OC (Newegg, related current offer) | $579.99 |
| Linux VPS with 8GB RAM and 4 CPU cores | $15-$25/month (as stated by source) |
Source prices are as stated by the source. Retailer offers: prices checked 2026-10-16, may change. Some links may be affiliate links; we may earn a commission at no extra cost to you. This never changes what we list.
Software stack
- Docker Compose
- Ollama
- n8n
- Caddy (HTTPS)
Models used
- Llama 3.2 3B (default)
- Llama 3.1 8B / Qwen 2.5 7B / Mistral 7B (options)
Good at
- Escalation to a human when the reply mentions 'human agent'
- Same workflow can answer email via IMAP
- Total $15-$25/month per the source
Weak at
- 3B model 'loses context in long back-and-forth chats'
- Can't look up a specific customer's order without extra database work
Rough total cost
$15-$25/month (VPS) per the source.
Getting started
- Run Ollama in Docker and pull llama3.2.
- Add n8n (with basic auth) and Caddy for HTTPS.
- Write an FAQ knowledge base (under ~2,000 words).
- Build the n8n workflow: webhook > prompt > Ollama > escalation check > reply.
- Embed the web widget or connect the support inbox.
Source & credit
Building a Customer Support Chatbot with Open Source AI: A Step-by-Step Guide — ARDOT Consulting blog
Summarized from the source; quotes are the source author's words. Check the original before buying.