Reproduce “Assessing the CAD Capabilities of General-Purpose AI Models: A Case Study of GPT-6 Astra” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Qiang Zou, Jiajun Shen, Zhijie Yang, Yuhao Wang, Shitianyi Leng, Dongshuo Yin

GPT-6 Astra assesses CAD understanding, generation, editing, and evaluation across six task types using a tool-enabled setup. In text-to-CAD, all 12 cases met listed requirement checks, but only 11 passed solid-validity checks. In parametric reconstruction, 52 of 60 eligible edits passed, while only four of 11 models passed all ten prescribed edits, limiting claims of dependable function.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.