Local-Models

• Local and frontier models — Run a local model for the routine 80% of the work and a frontier model for the hard 20% — how to wire each lane, and which work belongs on which.
• Ollama — Serves open models locally over a simple HTTP API — the engine behind the local lane, run once on a GPU machine and shared across the set.
• opencode — An open-source terminal coding agent. In this stack it drives the local lane — pointed at a local model, never at a paid subscription.

← All tags