Install Ollama

Speed up your workflows with free local AI models.

Why Ollama?

Ollama is an open-source platform for running and managing large language models locally on your machine. In Aetherflux, it powers AI agents, auto-completions, and runtime interpretations.

Ollama is an optional dependency. Aetherflux will run without it, but installing Ollama (and at least one model) is what unlocks the AI features described above. Presently, Ollama is the only supported way to leverage AI within Aetherflux.

Download the Ollama installer

Download the installer for your platform from ollama.com/download:

Windows 10 or later: OllamaSetup.exe
macOS 14 Sonoma or later: Ollama.dmg

If you would prefer, you can also install these using the appropriate terminal commands shown below.

Windows (PowerShell):

irm https://ollama.com/install.ps1 | iex

macOS / Linux:

curl -fsSL https://ollama.com/install.sh | sh

Install the application

Follow the instructions given by the installer of your choice. Once installation finishes, Ollama runs in the background and exposes a local service that Aetherflux can talk to. You do not need to keep a chat window open for Aetherflux to use it.

By default, Aetherflux expects Ollama on your local machine. If your workstation is short on memory, you can host Ollama on another machine (or a dedicated server) and change the Ollama instance address in Aetherflux to point at that remote host instead.

Pull a model

Ollama does not ship with model weights out of the box. After installing, open a terminal and pull a model of your choice. For example:

ollama pull llama3.2

Or start an interactive session (this will download the model if it is not already present):

ollama run llama3.2

Browse available models in the Ollama model library. Choose something that fits the memory and performance of your machine. Smaller models are faster and lighter; larger ones are more capable. Any models you install locally will automatically become accessible within Aetherflux.

Next steps

With Ollama running and at least one model pulled, open Aetherflux and start using AI-assisted workflows. If you have not installed the app yet, see install Aetherflux. Support for additional AI providers is planned. You can follow progress on our roadmap.