In this video, I’ll show you how to run OpenClaw locally using Ollama so you can build your own private AI agent system at home.

Instead of paying for expensive API tokens or relying on cloud services that can go down, you can run your own local AI infrastructure that works 24/7 on your own hardware.

I’ll walk through two setups:

• Beginner setup: Run OpenClaw and a local LLM on the same computer
• Advanced setup: Run OpenClaw on a small device (Jetson Nano / Raspberry Pi) and host the AI model on a separate machine

You’ll learn how to:

• Install Ollama
• Run local LLM models
• Install OpenClaw locally
• Connect OpenClaw to a remote Ollama server
• Choose the best local AI model for your hardware
• Set up static IP and Wake-on-LAN
• Build a home AI server that runs 24/7

By the end, you’ll have your own self-hosted AI agent system running locally.

No API costs.
No cloud dependency.
Your data stays private.

But I’ll also explain when running OpenClaw locally is NOT the best choice, and when you should just use a cloud setup instead.

⸻


00:00 Local OpenClaw Promise
00:30 Why Go Local
02:06 Cloud vs Local Explained
03:11 Two Setup Roadmap
04:22 Install Ollama Basics
04:49 Download and Test Models
05:42 Install OpenClaw via Ollama
06:31 Web UI and Gateway Token
07:37 Add Models to Config
10:47 Pick the Right LLM
14:00 Advanced Two Machine Setup
16:44 Ollama Server Networking
18:22 Static IP and Wake on LAN
20:14 Context Length and Power
22:59 Connect OpenClaw to Remote
26:17 Wrap Up and Recommendations

💬 Join my free AI community
(learn AI automation, AI agents, and workflows)
👉 https://nas.io/rumjahn

⸻

⚙️ Tools used in this video

OpenClaw
Ollama
LM Studio
Jetson Nano
Local LLM models (Qwen, Gemma)

⸻

If you’re interested in:

AI agents
Local LLMs
Self-hosted AI
Automation workflows
AI productivity

Subscribe for more videos on AI systems, automation, and building your own AI tools.

⸻

👍 If this video helped you, please like and subscribe — it really helps the channel.

YouTube

A step-by-step guide to running OpenClaw (an AI coding agent) locally using Ollama instead of cloud APIs. Covers two setups: a beginner single-machine setup and an advanced two-device setup using a Jetson Nano running OpenClaw 24/7 paired with an old gaming laptop as an Ollama LLM server. Topics include installing Ollama, choosing the right local model (balancing speed vs quality), configuring networking with static IPs, enabling Wake-on-LAN, setting context length, and connecting OpenClaw to a remote Ollama server. Key motivations are cost savings, privacy, and reliability when cloud providers go down.

Local OpenClaw & Ollama in 27 minutes