Installation

There are several ways to get llama.cpp on your machine:

• Recommended (Mac/Windows): one click install the desktop apps
• Recommended (others): one-line install command
• Install with a package manager
• Download prebuilt binaries from the releases page
• Run with Docker
• Build from source

All of them give you the same set of tools (llama cli, llama serve and others).

One-click install (recommended)

If you intend to use LLMs on Mac or Windows, you can download the app from the main page, which launches right after installation.

One-line install (recommended)

This is the easiest way to get started with llama.cpp. The following command detects your platform, fetches the latest version of the llama binary and installs it.

curl -LsSf https://llama.app/install.sh | sh

To pair with a coding agent, install with a package manager or build yourself, please follow the instructions in https://llama.app.

Verify the installation

If you have downloaded the Llama app, it should show up on your tool bar.

If you want to use the CLI, or you are on a platform that doesn’t support the app, type:

llama cli --version

If this prints the version and build info, you are ready to go. Continue with the Quickstart to download and run your first model.