Home Assistant Ollama-HA App

Run Ollama locally on your Home Assistant instance.

Installation Instructions

Tip

For detailed installation instructions of the App and examples, download my PDF: https://automatelike.pro/ollama-ha

Warning

Large Download Warning: The app image is over 3.7GB. Installation can take 25-30 minutes or more on a Raspberry Pi 5, depending on your internet speed and SD card performance. Please be patient and ensure you have sufficient disk space.

1. Add Repository:
   • Go to Settings > Apps > App Store.
   • Click the three dots in the top right > Repositories.
   • Add the URL of this repository: https://github.com/peyanski/ollama-ha.
2. Install App:
   • Scroll down or search for "Ollama".
   • Click Install. (This is the long step!)
3. Start App:
   • Once installed, toggle "Start on boot" and click Start.
   • Check the "Log" tab to ensure the server starts listening on port 11434.

Configuration

GPU Support

By default, GPU support is disabled to ensure compatibility. If you have a supported GPU passed through to Home Assistant:

1. Go to the Configuration tab.
2. Toggle gpu to true.
3. Restart the app.

Manage Models (List & Delete)

To free up disk space or check what is installed, use the command line.

Prerequisites

1. Install the Advanced SSH & Web Terminal app.
2. Start it and open the terminal.

Accessing the Ollama Terminal

1. Find Container Name:

   docker ps | grep ollama-ha

   (Look for a name like addon_ad7c61ed_ollama-ha).

2. Enter Container:

   docker exec -it addon_ad7c61ed_ollama-ha bash

   (Replace addon_ad7c61ed_ollama-ha with your actual container name).

List & Delete

Once inside the container:

• List Installed Models: Check which models are using space.

  ollama list

• Delete a Model: Remove a model by name.

  ollama rm llama3.2