Ollama Stack

Deploy a local LLM stack for offline and privacy-first workflows.

When to Use This Skill

Use this skill when:

• Setting up private/local LLM inference for development
• Building air-gapped AI environments
• Running models on personal hardware (Mac, Linux, Windows with GPU)
• Creating team-shared inference endpoints
• Prototyping before committing to cloud LLM APIs

Deployment boundaries

Pin reviewed container images and model revisions before use; table sizes/VRAM are illustrative and context/concurrency change requirements. A local endpoint does not guarantee offline operation: model downloads, WebUI features and proxy backends can make network calls. Confirm mounted data, authentication and network policy; the nginx IP restriction below is not user authentication. Container OLLAMA_HOST=0.0.0.0 is for the private Compose network while published host ports bind to loopback. See Ollama networking guidance.

Prerequisites