Run a private local AI with Ollama (free and offline)

You can run an AI model directly on your own computer with Ollama: private, offline, and free. Nothing leaves your machine, there is no API key to buy, and it keeps working with no internet connection. The Local AIs section of the Connectors tab walks you through the whole thing.

Where this lives

Open the Connectors tab (right pane More menu, then Connectors) and scroll to the Local AIs section.

Step 1: Install Ollama

1. Click Get Ollama. It opens ollama.com/download in your browser.
2. Download the installer for your system and run it like any other app.
3. Come back to the Local AIs section once it is installed.

Ollama is the small engine that runs models on your machine. You only install it once.

Step 2: Download a model

With Ollama installed, the section shows two lists:

• Already-installed models, each with its size on disk, so you always know what you have and how much space it takes.
• A catalog of suggested models, each row with a Download button.

1. Pick a model from the catalog and click Download.
2. Watch the live progress line on the row. Models are large files, so this can take a while on slower connections - the progress line always shows where things stand.
3. Changed your mind mid-download? Click Stop to cancel. You can start again later.

Once the download finishes, the model appears in your installed list with its size, ready to use.

Why bother with a local AI?

• Private - the model runs on your machine, so what you feed it stays with you.
• Free - no API key, no per-use cost. The only price is disk space and some patience during the download.
• Offline - once a model is installed it works without an internet connection, which is handy backstage, in transit, or anywhere the venue Wi-Fi is a rumor.

Practical tips

• Check the sizes in the installed list before downloading several models - they add up on disk.
• If a download is crawling, Stop it and retry when you are on a faster connection.
• Local models are a great private option, though the cloud providers you can connect on this same tab (Claude, ChatGPT, or Gemini) remain available whenever you want them.

Still stuck? Submit a ticket from this help center and we'll take it from there.