- Posted on
- Featured Image
A practical, bash-first guide to running multiple local LLM micro-services on Linux for privacy, cost control, and speed: build llama.cpp, run two GGUF servers on separate ports, cap CPU/GPU/memory with taskset/nice/cgroups, put a single NGINX front door with path-based routing, manage with systemd user services, monitor usage, and optionally use Ollama for a simpler multi-model setup, plus tips and real-world patterns.