<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>GPU on Manvendra Rajpoot</title>
    <link>https://blog.rajpoot.dev/tags/gpu/</link>
    <description>Recent content in GPU on Manvendra Rajpoot</description>
    <image>
      <title>Manvendra Rajpoot</title>
      <url>https://blog.rajpoot.dev/img/personal/cover.png</url>
      <link>https://blog.rajpoot.dev/img/personal/cover.png</link>
    </image>
    <generator>Hugo</generator>
    <language>en</language>
    <copyright>Manvendra Rajpoot</copyright>
    <lastBuildDate>Sun, 17 May 2026 17:50:46 +0530</lastBuildDate>
    <atom:link href="https://blog.rajpoot.dev/tags/gpu/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Self-Hosted LLMs in 2026 — Ollama, vLLM, and When to Skip the API</title>
      <link>https://blog.rajpoot.dev/posts/ai/self-hosted-llms-vllm-ollama-2026/</link>
      <pubDate>Tue, 28 Apr 2026 20:50:00 +0530</pubDate>
      <guid>https://blog.rajpoot.dev/posts/ai/self-hosted-llms-vllm-ollama-2026/</guid>
      <description>When to self-host LLMs in 2026 — Ollama for dev, vLLM and SGLang for production, model choice, hardware sizing, and the latency/cost tradeoffs vs hosted APIs.</description>
    </item>
  </channel>
</rss>
