How to Run AI on Your Own Computer: Ollama Guide for Australians
Step-by-step guide to running open-source AI models locally on your PC or Mac. Free, private, and no internet required.
AINow Take
Running AI locally is easier than you think. If you have a computer made in the last 3 years, you can run a useful AI model right now for free. Start with Llama 3.2 (8B) — it runs on most laptops.
Running AI on your own computer gives you privacy, no subscription costs, and unlimited usage. Here’s exactly how to do it.
What You Need
| Hardware | Minimum | Recommended |
|---|---|---|
| RAM | 8GB | 16GB+ |
| Storage | 10GB free | 50GB+ free |
| GPU (optional) | None | NVIDIA 6GB+ VRAM |
| Internet | For download only | For download only |
Yes, you can run AI on a laptop without a GPU. It will be slower, but it works.
Step 1: Install Ollama
Ollama is the easiest way to run local AI models.
Download: ollama.com — available for macOS, Linux, Windows
Install: Download and run the installer. That’s it.
Step 2: Run Your First Model
Open a terminal (Command Prompt on Windows) and run:
ollama run llama3.2This downloads and starts Llama 3.2 (3B model, ~2GB). Wait for the download to finish, then type your questions.
First-time download takes 2-10 minutes depending on your internet.
Step 3: Try Different Models
| Model | Size | Quality | Hardware |
|---|---|---|---|
llama3.2 (3B) |
~2GB | Good | Any computer |
llama3.1 (8B) |
~4.7GB | Very good | 8GB+ RAM |
mistral (7B) |
~4.1GB | Very good | 8GB+ RAM |
phi3 (3.8B) |
~2.3GB | Good | Any computer |
llama3 (70B) |
~40GB | Excellent | 32GB+ RAM / GPU |
Start with llama3.2 (3B). It runs on almost anything and is surprisingly capable.
Step 4: Add a Web Interface
Ollama is command-line only by default. For a ChatGPT-like experience:
Open WebUI (Recommended)
- Install Python
- Run:
pip install open-webui - Run:
open-webui serve - Open http://localhost:8080
Or use: LM Studio — a desktop app with a built-in GUI for local models.
Step 5: Use AI in Other Apps
Once Ollama is running, it acts as a local API server. Other apps can use it:
- Continue.dev — VS Code extension (AI coding assistant)
- Enchanted — macOS ChatGPT client that works with Ollama
- TypingMind — Custom UI that supports local models
Australian-Specific Tips
Privacy
Running AI locally means your data never touches the internet. This is important for:
- Legal document review
- Medical information
- Financial data
- Client communications covered by Australian privacy law
Cost Comparison
| Setup | One-time Cost | Monthly | 12-month Total |
|---|---|---|---|
| Local Llama 3.2 | $0 (existing PC) | $0 (electricity ~$2) | ~$24 |
| ChatGPT Plus | $0 | ~$31 AUD | ~$372 |
| ChatGPT Pro | $0 | ~$310 AUD | ~$3,720 |
Local AI saves $348-3,696 AUD per year.
Internet Outages
Australia still has internet reliability issues in some areas. Local AI works completely offline.
What You Can’t Do (Yet)
Local models are amazing but not perfect:
- Slower than cloud AI (especially on laptops)
- Smaller knowledge cutoffs
- No built-in web search
- Less capable at complex reasoning
But for daily tasks — writing, brainstorming, analysis, coding help — local AI is genuinely useful.
Quick Start (5 Minutes)
# 1. Download Ollama from ollama.com
# 2. Open terminal and run:
ollama run llama3.2
# 3. Start chattingThat’s it. You now have AI on your computer, free forever.