Skip to main content

llama (server-side) CLI Reference

The ogx CLI tool helps you set up and use the OGX. The CLI is available on your path after installing the ogx package.

Installation​

You have two ways to install OGX:

  1. Install as a package: You can install the repository directly from PyPI by running the following command:

    pip install ogx

  2. Install from source: If you prefer to install from the source code, follow these steps:

     mkdir -p ~/local
     cd ~/local
     git clone git@github.com:ogx-ai/ogx.git

     uv venv myenv --python 3.12
     source myenv/bin/activate  # On Windows: myenv\Scripts\activate

     cd ogx
     pip install -e .

ogx subcommands​

  1. stack: Allows you to build a stack using the ogx distribution and run a OGX server. You can read more about how to build a OGX distribution in the Build your own Distribution documentation.
  2. connect: Connect third-party tools to the running OGX server. Supports claude, opencode, and codex.

For downloading models, we recommend using the Hugging Face CLI. See Downloading models for more information.

Sample Usage​

llama --help
usage: ogx [-h] {stack,connect} ...

Welcome to the OGX CLI

options:
  -h, --help      show this help message and exit

subcommands:
  {stack,connect}

  stack                 Operations for the OGX / Distributions
  connect               Connect third-party tools to the running OGX server

Downloading models​

You first need to have models downloaded locally. We recommend using the Hugging Face CLI to download models.

First, install the Hugging Face CLI:

pip install huggingface_hub[cli]

Then authenticate and download models:

# Authenticate with Hugging Face
huggingface-cli login

# Download a model
huggingface-cli download meta-llama/Llama-3.2-3B-Instruct --local-dir ~/.ogx/Llama-3.2-3B-Instruct

List the downloaded models​

To list the downloaded models, you can use the Hugging Face CLI:

# List all downloaded models in your local cache
huggingface-cli scan-cache