By the end of this guide, you will have an AI model running on your own computer: no account, no subscription, and nothing you type is sent to the internet.
What you need
- A computer with 8 GB of RAM or more (16 GB is more comfortable).
- About 5 GB of free disk space for a first model.
- An internet connection for the download only. After that, it works offline.
Tip
Not sure what your computer can handle? Read Choose the right local model for your computer first, or just follow this guide: the model we use is small.
Step 1: Install Ollama
Windows
- Go to ollama.com/download and download the Windows installer.
- Run
OllamaSetup.exeand follow the steps. - Ollama starts in the background (llama icon near the clock).
macOS
- Download Ollama for macOS from the same page.
- Open the file and move Ollama to your Applications folder, then launch it.
Linux
Open a terminal and run:
curl -fsSL https://ollama.com/install.sh | shStep 2: Open a terminal
- Windows: press the Windows key, type
PowerShell, press Enter. - macOS: press Cmd + Space, type
Terminal, press Enter. - Linux: open your terminal app.
Check that Ollama is installed:
ollama --versionYou should see a version number.
Step 3: Download and run a model
Type:
ollama run llama3.2The first time, Ollama downloads the model (about 2 GB). When you see >>>, the AI is ready. Ask it something:
Explain in 3 simple sentences what a local AI model is.To leave the chat, type /bye. Next time, ollama run llama3.2 starts instantly because the model is already on your disk.
Step 4: Try other models
Each model has strengths. Browse the Ollama library and try a few. For example:
ollama run gemma3
ollama run qwen3See and clean up what you downloaded:
ollama list
ollama rm gemma3Note
Model names and versions change often. If a command says the model is not found, check its exact name in the library.
Step 5: Add a nicer interface (optional)
The terminal works, but you can also chat in a window:
- Open WebUI: a ChatGPT-like web page on top of Ollama. Follow this guide.
- Many apps and editor extensions can use Ollama through its local API at
http://localhost:11434.
Troubleshooting
| Problem | Solution |
|---|---|
ollama: command not found | Close and reopen the terminal; on Windows/macOS make sure the Ollama app is running |
| Answers are very slow | Use a smaller model (1–4B parameters) and close heavy apps |
| Download fails | Check your connection and free disk space, then run the command again |
| The computer freezes | The model is too large for your RAM; remove it with ollama rm and pick a smaller one |
Last checked: mid-2026. Menus and button names in AI apps change often, so if something looks different, look for the closest match.
Comments 0
Share tips, prompts that worked for you, or questions.
No comments yet. Be the first to share a tip.
Log in or create a free account to comment.