Lakehouse Software
Home / Local AI Lab / 02

Local model runtime

Install Ollama

Install the engine that will run the model locally and expose it to our assistant through a localhost API.

Chapter 02/10Time 10–20 minCost $0 software

Why Ollama?

Ollama provides a simple local model runtime on Windows, macOS, and Linux and exposes a local HTTP API that our controller can call later. That keeps the tutorial focused on building the assistant rather than implementing model inference.

Install Ollama

WindowsmacOSLinux

Windows

  1. Download Ollama for Windows.Use the official Ollama download page and run the installer.
  2. Open PowerShell.Verify the command is available.
ollama --version

macOS

  1. Download Ollama for macOS.Install the app from the official download page.
  2. Open Terminal.Verify the CLI.
ollama --version

Linux

The current official one-line installer is:

curl -fsSL https://ollama.com/install.sh | sh

Then verify:

ollama --version

Confirm the local service

Ollama normally exposes its API only on the local machine at port 11434. We will use that API in Chapter 5.

curl http://localhost:11434/api/tags
If curl is unavailable on your Windows setup

Do not get stuck here. If ollama --version works, continue. We will test the API from Python later.

Checkpoint

ollama --version returns a version and the Ollama application/service is running.

Official references

Ollama quickstart · Linux install

Finished this chapter?
Your progress is stored only in this browser.