<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>VoidSource Articles</title>
    <link>https://voidsource.dev/en/articles</link>
    <description>Evergreen analysis, benchmark explainers, and field guides for building with AI systems.</description>
    <language>en</language>
    <lastBuildDate>Thu, 26 Mar 2026 00:00:00 GMT</lastBuildDate>
    <atom:link href="https://voidsource.dev/en/articles/feed" rel="self" type="application/rss+xml"/>
    <image>
      <url>https://voidsource.dev/logo.png</url>
      <title>VoidSource</title>
      <link>https://voidsource.dev</link>
    </image>
    
    <item>
      <title><![CDATA[We Benchmarked 7 OCR Models So You Don&apos;t Have To]]></title>
      <link>https://voidsource.dev/en/articles/olmocr-bench-reproduction</link>
      <guid isPermaLink="true">https://voidsource.dev/en/articles/olmocr-bench-reproduction</guid>
      <description><![CDATA[Results from our own olmOCR-bench runs across OCR-specific models, general-purpose VLMs, and Tesseract. The main lesson: evaluation methodology changes scores more than most leaderboard readers realize.]]></description>
      <pubDate>Thu, 26 Mar 2026 00:00:00 GMT</pubDate>
      <author>VoidSource</author>
      <category>ocr</category>
      <category>document-ai</category>
      <category>benchmarks</category>
      <category>evaluation</category>
      <category>olmocr</category>
      <category>vlm</category>
    </item>
    <item>
      <title><![CDATA[Claude Code is the New Cursor (and the Cycle Never Ends)]]></title>
      <link>https://voidsource.dev/en/articles/claude-code</link>
      <guid isPermaLink="true">https://voidsource.dev/en/articles/claude-code</guid>
      <description><![CDATA[Everyone&apos;s screaming about Claude Code. A year ago, they screamed about Cursor. Before that, Copilot. The pattern is more interesting than the product, and reveals something uncomfortable about how we adopt tools.]]></description>
      <pubDate>Sun, 18 Jan 2026 00:00:00 GMT</pubDate>
      <author>VoidSource</author>
      <category>claude-code</category>
      <category>coding-agents</category>
      <category>ai-tools</category>
      <category>meta</category>
      <category>hype-cycles</category>
    </item>
    <item>
      <title><![CDATA[When to Build Your Own AI Agent (and When Claude Code Is Enough)]]></title>
      <link>https://voidsource.dev/en/articles/building-ai-agents-anthropic-sdk</link>
      <guid isPermaLink="true">https://voidsource.dev/en/articles/building-ai-agents-anthropic-sdk</guid>
      <description><![CDATA[When should you build your own AI agent instead of using Claude Code? A practical guide to the Anthropic Agent SDK, with mental models for understanding when and why to go beyond interactive AI.]]></description>
      <pubDate>Sat, 17 Jan 2026 00:00:00 GMT</pubDate>
      <author>VoidSource</author>
      <category>ai-agents</category>
      <category>llm</category>
      <category>sdk</category>
      <category>claude</category>
      <category>anthropic</category>
      <category>developer-tools</category>
      <category>python</category>
      <category>typescript</category>
    </item>
    <item>
      <title><![CDATA[Which Document AI Benchmarks Actually Matter?]]></title>
      <link>https://voidsource.dev/en/articles/document-processing-benchmark-atlas</link>
      <guid isPermaLink="true">https://voidsource.dev/en/articles/document-processing-benchmark-atlas</guid>
      <description><![CDATA[A clear guide to document processing benchmarks. Understand what each evaluation actually measures, when to use it, and what the scores mean.]]></description>
      <pubDate>Thu, 15 Jan 2026 00:00:00 GMT</pubDate>
      <author>VoidSource</author>
      <category>benchmarks</category>
      <category>evaluation</category>
      <category>ocr</category>
      <category>document-ai</category>
      <category>tables</category>
      <category>forms</category>
      <category>docvqa</category>
    </item>
    <item>
      <title><![CDATA[Docling: What IBM&apos;s 37K-Star Toolkit Actually Does]]></title>
      <link>https://voidsource.dev/en/articles/docling</link>
      <guid isPermaLink="true">https://voidsource.dev/en/articles/docling</guid>
      <description><![CDATA[IBM&apos;s Docling toolkit has 37k+ GitHub stars. But what does it actually do? A clear explanation of when document conversion needs more than OCR, and whether Docling solves your specific problem.]]></description>
      <pubDate>Tue, 28 Jan 2025 00:00:00 GMT</pubDate>
      <author>VoidSource</author>
      <category>ocr</category>
      <category>idp</category>
      <category>rag</category>
      <category>pdf</category>
    </item>
    <item>
      <title><![CDATA[Document Processing Benchmarks Are a Mess. Here&apos;s How to Read Them.]]></title>
      <link>https://voidsource.dev/en/articles/document-processing-benchmarks</link>
      <guid isPermaLink="true">https://voidsource.dev/en/articles/document-processing-benchmarks</guid>
      <description><![CDATA[A practical guide to evaluating document AI. Mental models for understanding OCR, table extraction, and document understanding benchmarks without drowning in metrics.]]></description>
      <pubDate>Tue, 14 Jan 2025 00:00:00 GMT</pubDate>
      <author>VoidSource</author>
      <category>ocr</category>
      <category>idp</category>
      <category>benchmarks</category>
      <category>evaluation</category>
      <category>rag</category>
    </item>
  </channel>
</rss>