What this does

This guide installs the Ollama runtime on a macOS host using the official installer package. After completion, the ollama daemon runs in the background and accepts inference requests on port 11434.

Steps

Download the macOS installer from the Ollama website. The package includes the CLI binary and a LaunchAgent for automatic startup.
```
curl -fsSL https://ollama.com/install.sh -o install.sh
```
Expected output: File download progress ending in a saved install.sh script.
Run the installer script. The script places the binary at /usr/local/bin/ollama and registers the LaunchAgent.
```
chmod +x install.sh
./install.sh
```
Expected output: Ollama has been installed. Run 'ollama' to start.
Start the Ollama service interactively or rely on automatic startup. On macOS, a LaunchAgent handles background operation.
```
launchctl start ai.ollama.ollama
```
Expected output: No output indicates success.

Record the local run evidence. Save the exact command, runtime or package version, model name if applicable, and observed output so the result can be reproduced later.

Verification

ollama list
# Expected: NAME   ID   SIZE   MODIFIED  (empty table is normal)

Common failures

Installation hangs on download — Slow or interrupted connection. Retry the download or use a direct binary install instead.
launchctl reports service not found — The LaunchAgent may not have registered correctly. Re-run the install script or start manually with ollama serve.
Binary not found in new terminal — PATH issue. Verify /usr/local/bin is in PATH with echo $PATH | tr ':' '\n' | grep local.
Apple Silicon (M-series) performance warnings — Ollama uses native ARM64 binaries on Apple Silicon automatically. No action needed.
Port 11434 already in use — Another process occupies the port. Identify it with lsof -i :11434 and stop or reassign the conflicting service.

How to install Ollama on macOS

What this does

Steps

Verification

Common failures

Related guides