Complete procedures with a way back.

Build a repeatable capacity-and-quality matrix for local GGUF models instead of choosing from a model leaderboard or a vendor logo. Inventory the real machine, calculate a safe memory envelope, pin Qwen3 4B, 8B, and 14B candidates, run llama.cpp b9637 with the appropriate CPU, Metal, CUDA, ROCm, or Vulkan backend, and select the smallest candidate that passes the workload at the required concurrency.

Select a licensed, checksum-backed GGUF and measured quantization, build llama.cpp from a pinned commit, size CPU/GPU resources, validate metadata and quality, isolate llama-server behind a trusted proxy, benchmark real workloads, and roll back the complete binary/model bundle.