Server and Model Inventory

One background service, several clients

Ollama is a tool that downloads, stores, and runs language models. It starts a background program called a daemon. That daemon provides a local Hypertext Transfer Protocol (HTTP) service. The ollama command-line interface (CLI), the curl web-request tool, and Python programs are all clients of it. ollama list shows installed models; ollama show MODEL shows descriptive data called metadata; ollama ps shows models loaded in memory now.

ollama run qwen2.5:0.5b opens an interactive conversation. A one-shot form such as ollama run qwen2.5:0.5b 'Return only the word READY' is easier to script and test.

Note: The lab image already contains its models. Do not run ollama pull in a session. The lab has no internet access, and downloading at runtime would make sessions slow and hard to reproduce.