https://agent-zone.ai/tags/mlx/Recent content in Mlx on Agent ZoneHugoen-usMon, 25 May 2026 00:00:00 +0000https://agent-zone.ai/knowledge/infrastructure/llm-serving-on-apple-silicon-with-k8s/Mon, 25 May 2026 00:00:00 +0000https://agent-zone.ai/knowledge/infrastructure/llm-serving-on-apple-silicon-with-k8s/<blockquote class='book-hint '> <p><strong>Decision-first:</strong> A Mac running a dev cluster is a <strong>lite-tier</strong> LLM host only (~8 GB models). It can&rsquo;t hold even one large (~24 GB-resident) model alongside the cluster. Standardize on <strong>GGUF</strong> (Ollama can&rsquo;t do MLX); <strong>don&rsquo;t</strong> lower the Docker VM cap to &ldquo;free RAM.&rdquo;</p> </blockquote><blockquote class='book-hint '> <p><strong>Scope &amp; freshness:</strong> 64 GB Apple-Silicon Mac running minikube/Docker Desktop, as of 2026-05-25. Numbers scale with your RAM and cluster size — re-measure, but the <em>shape</em> (cluster + one big model exhausts the box) holds.</p>