Ollama
Runs large language models on your own server and serves them through an API that other apps use. It has no web interface of its own; Open WebUI is one app that talks to it. After install, pull and run a model with docker exec -it <container> ollama run llama3.2, or from an app such as Open WebUI. Models are several gigabytes each. The API has no login: anyone who can reach its port can run, pull and delete models. Keep it on your home network and never forward it from your router. Without a GPU, models run on the CPU.
- Documentation
- https://docs.ollama.com/docker
- Template revision
- 3
Install Ollama
In Hoserva, open Apps → Catalog, select Ollama and fill in the install form. This page only lists the app; nothing here installs it.