<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>CAD-Coder: An Open-Source Vision-Language Model for Computer-Aided Design Code Generation · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2505.14646/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/2505.14646/feed.xml" rel="self" type="application/rss+xml" /><item><title>CAD-Coder: An Open-Source Vision-Language Model for Computer-Aided Design Code Generation — mentioned by KerrickStaley</title><link>https://news.ycombinator.com/item?id=47213509</link><guid isPermaLink="false">urn:hackerpapers:mention:2505.14646:47213509</guid><pubDate>Mon, 02 Mar 2026 03:28:29 GMT</pubDate><description>&lt;p&gt;I recently designed an eval to see if LLMs can produce usable CAD models: https://kerrickstaley.com/2026/02/22/can-frontier-llms-solve... Claude 4.6 Opus and Gemini 3.1 Pro can to some degree, although the 3D models they produce are often deficient in some way that my eval didn&amp;#x27;t capture. My eval used OpenSCAD simply due to familiarity and not having time to experiment with build123d/CadQuery. There is an academic paper where they were successful at fine-tuning a small VLM to do CadQuery: https://arxiv.org/pdf/2505.14646&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2505.14646/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item></channel></rss>
