Unified Controllable and Faithful Text-to-CAD Generation with LLMs

AI & Machine Learning, Computer Science > Computation and Language(arxiv.org)view on HackerNews

text-to-CADlarge language modelsprogressive refinementcontrollablefaithfulCAD generationreinforcement learning

Author: PaulHoule

Date: 6/9/2026

Article Summary:

This paper proposes PR-CAD, a progressive refinement framework for unified controllable and faithful text-to-CAD generation with large language models.