Ever wished you could use ChatGPT but don’t want your sensitive data floating around on someone else’s servers? Or maybe your organization blocks cloud AI tools for privacy reasons?
Good news you can run powerful language models entirely on your own laptop. For free. Completely private. No internet required after setup.
Let me show you how.
What Exactly is Ollama?
Ollama is an open-source tool that lets you download and run large language models (LLMs) directly on your local machine no API calls, no data leaving your computer, no token limits.
Think of it as having your own private ChatGPT living inside your laptop.
Step 1: Download Ollama
Head over to ollama.com
You’ll see a download button. Click it and grab the version for your operating system:
- macOS
- Linux
- Windows
I’m on a Mac, so I downloaded the macOS version.
Step 2: Install Ollama
Once the file downloads:
- Open the installer
- Drag the Ollama icon into your Applications folder
- Open Finder, search for “Ollama” and launch it
- Approve the security prompt (it’s safe, I promise!)
- Enter your password when asked
You’ll see a small llama icon appear in your menu bar. That’s Ollama running quietly in the background.
Step 3: Find a Model to Download
Ollama has a library of models you can pull locally.
Visit ollama.com/models and browse around. You’ll see:
- Model sizes (anywhere from a few hundred MB to 200+ GB)
- Parameter counts (more parameters = smarter but heavier)
- Different model families (Llama, Mistral, Gemma, etc.)
For testing, I’d recommend starting small. I tried “Tiny Llama” first (spoiler: it was hilarious and weird). Then I moved to Llama 3.2 1B which is much more capable.
Step 4: Open Terminal and Pull Your First Model
Open your terminal and type:
ollama run llama3.2:1bThat’s it. The model starts downloading automatically.
You’ll see progress bars and once it’s done boom. You’re talking to an AI running entirely on your machine.
Step 5: Start Chatting
Once the download completes, you can start typing right in the terminal.
Ask it anything. Write code. Get recipe ideas. Whatever you want.
The responses come fast because there’s no network lag. It’s all happening inside your computer.
Two Ways to Chat
Option A: Terminal
Just keep typing in the terminal window. When you’re done, type /bye to exit.
Option B: Ollama App
Open the Ollama app from your menu bar. You’ll see a dropdown with all your downloaded models. Pick one and start a conversation right there nicer interface, same local AI.
Why I Love This Setup
- Complete privacy—my data never leaves my laptop
- Zero cost—no API bills, no subscription
- No limits—chat all day if you want
- Works offline—take it on a plane, to a cafe, wherever
A Quick Warning About Model Sizes
Some models are massive (200GB+). Make sure you have enough disk space before downloading the big ones. Start small, experiment, then level up when you’re ready.
That’s It!
You now have a private, offline, completely free AI assistant running on your own machine.
If you hit any snags or have questions, drop them in the comments below. I read every single one.
Until next time happy local prompting!
