Free AI for n8n: How To Connect Ollama to n8n Workflows – No API Costs! (Part 2)

Description

In this Part 2, learn how to connect n8n to shared Docker Network and how to use Ollama AI Models in n8n automation workflows with no API Costs, Rate Limits and 100% free (excluding the cost of monthly Hostinger VPS Plan).

What You Will Learn (How):

• Inspect Docker Network
• Terminal Tips
• Edit n8n docker-compose.yml file
• 3-Step Nano Editor Process (Write out, Commit, Exit)
• Stop & Restart n8n Docker container
• Verify that n8n is connected to the shared network
• Setup Ollama Chat Model Node credentials (internal URL)
• Test Ollama Models
• How To Pull, Remove & List Ollama Models

Tutorials:
(Part 1: Ollama + Docker Network): https://youtu.be/9-u3BGl0ubU
(Part 2: Ollama + N8N Workflows): https://youtu.be/5qFNxvj59hw

Cheatsheet (Snippets) on GitHub:
https://dub.sh/ollama_1 (Part 1)
https://dub.sh/ollama_2 (Part 2)

Resources Mentioned:

• Hostinger VPS (https://mikemurphy.co/vps)
• Ollama https://olama.com
• N8N (Self-hosted)
• Shared Docker Network (App.net)
• docker-compose.yml file
• Nano text editor
• GitHub repo cheatsheet https://dub.sh/ollama_1

Ollama Models:
llama 3.2:latest (https://ollama.com/library/llama3.2:latest)
qwen 2.5:0.5b (https://ollama.com/library/qwen2.5:0.5b)
gemma2:2b (https://ollama.com/library/gemma2:2b)

Chapters:
00:00 Intro: About The Tutorial
00:18 What You Will Learn
00:42 Recap: Part 1
01:20 Inspect Docker Network
02:04 Containers (Connected Apps)
02:45 Change Directory (n8n)
03:20 Root Directory
04:11 Edit docker-compose.yml
04:58 Nano Editor Navigation
06:09 Add Networks To YAML file
07:43 Write Out Nano File
09:06 Restart n8n Container
10:46 YAML Editor (VPS Dashboard)
11:22 Fix syntax error in .yaml editor
12:10 Confirm n8n is connect to Docker Network (app-net)
14:02 Docker Container Ports
14:20 Test Ollama in n8n
14:44 Create new n8n workflow
15:18 Edit Fields (Set)
15:49 Ollama Chat Model Node
16:18 Docker Container Internal URLs
17:31 Base URL (Ollama Node)
18:39 Select Ollama Model
19:19 Customize LLM Chain
19:44 Execute Workflow
19:58 On the fence about Ollama?
20:50 Try, try again!! (don’t quit on models!)
21:21 Test different Ollama Models
21:59 Chat with Ollama Models
23:07 Pull & Remove Ollama Models
23:55 Test New Model in n8n
24:31 List all Models Installed
24:47 GitHub Repo (Cheatsheets)

Step-by-Step (Ollama Setup on VPS):

1. Sign in to the Hostinger dashboard and launch the terminal.
2. Check connected containers to the App.net network
3. Change directory (CD) into the N8N folder if it was moved from the default root directory.
4. Open the existing docker-compose.yml file for n8n
5. Add network settings in docker-compose.yaml
6. Save the file using Ctrl+O, tap Return, and exit using Ctrl+X.
7. Stop and remove the N8N Docker container using docker compose down.
8. Restart and recreate the N8N container
9. Confirm N8N is connected to App.net
10. In the N8N workflow editor, add the Ollama Chat Model node.
11. Create a new credential for Ollama
12. Set Base URL to the internal address: http://olama:11434.
13. Select a model and configure the Basic LM Chain node
14. Execute the workflow to confirm successful communication between N8N and Ollama.
15. Pull new models, remove unwanted, list all Models