https://cruver.ai/categories/ai/ Recent content in Ai on Hullabalooing Hugo en Tue, 20 Jan 2026 00:00:00 +0000 https://cruver.ai/gpu-ai/posts/claude-code-local-llms-1768966311/ Tue, 20 Jan 2026 00:00:00 +0000 https://cruver.ai/gpu-ai/posts/claude-code-local-llms-1768966311/ <p>Every query to Claude Code means sending my source code to Anthropic&rsquo;s servers. For proprietary codebases, that&rsquo;s a non-starter. With vLLM and LiteLLM, I can point Claude Code at my own hardware - keeping my code on my network while maintaining the same workflow.</p> <h2 id="the-architecture">The Architecture<a class="anchor" href="#the-architecture">#</a></h2> <p>The trick is that Claude Code expects the Anthropic Messages API, but local inference servers speak OpenAI&rsquo;s API format. LiteLLM bridges this gap. It accepts Anthropic-formatted requests and translates them to OpenAI format for my local vLLM instance.</p> https://cruver.ai/gpu-ai/posts/an-affordable-ai-server-1768704467/ Sat, 17 Jan 2026 00:00:00 -0700 https://cruver.ai/gpu-ai/posts/an-affordable-ai-server-1768704467/ <p>Two AMD MI60s from eBay cost me about $1,000 total and gave me 64GB of VRAM. That&rsquo;s enough to run Llama 3.3 70B at home with a 32K context window.</p> <p>When I started looking into running large language models locally, the obvious limiting factor was VRAM. Consumer GPUs top out at 24GB, and that&rsquo;s on an RTX 4090 at the high end. I wanted to run 70B parameter models locally, on hardware I own.</p>