PR-CAD: Progressive Refinement for Text-to-CAD Generation with LLMs
Original: Unified Controllable and Faithful Text-to-CAD Generation with LLMs
PR-CAD unifies text-to-CAD generation and editing through progressive LLM-driven refinement.
This arXiv paper introduces PR-CAD, a framework for controllable and faithful text-to-CAD generation with large language models. It treats CAD creation and editing as one progressive refinement process rather than separate tasks. The authors curate an interaction dataset and report state-of-the-art controllability and faithfulness on public benchmarks.
This paper focuses on Text-to-CAD, which translates natural language descriptions into editable CAD models. Traditional CAD modeling requires extensive manual operations and specialized knowledge, but recent advances in LLMs have led researchers to explore generating CAD using text. However, the authors point out that existing methods often treat "generating from zero" and "editing later" as separate tasks, which limits practical usability, since the real design process is usually not completed all at once but involves continuously creating, adjusting, and revising parameters and local geometry.
Free shows the 3-line summary; Pro unlocks the full deep summary (~300 words) so you never have to click through.
See Pro plans →Want the original English / full article?
Read on Hacker News (AI keywords) →Summaries are AI-generated; the original article is authoritative.