Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Text-to-CadQuery: A New Paradigm for CAD Generation with Scalable Large Model Capabilities

Haoyang Xie, Feng Ju

TL;DR

This work targets the barrier to CAD generation by directly producing executable CadQuery code from natural language, bypassing intermediate command sequences. It introduces Text-to-CadQuery, a large-scale dataset of ~170k text–CadQuery pairs and a multi-model fine-tuning study that demonstrates a clear scaling trend: larger pretrained decoders yield better geometric fidelity and executability in CadQuery scripts. By leveraging a CadQuery-focused output and a data-annotated refinement loop, the approach achieves state-of-the-art results on standard CAD-geometric metrics, while enabling faster, code-based 3D generation without training from scratch. The findings suggest significant practical impact for scalable, language-driven CAD, with potential to extend to additional inputs and even larger models as data and compute resources grow.

Abstract

Computer-aided design (CAD) is fundamental to modern engineering and manufacturing, but creating CAD models still requires expert knowledge and specialized software. Recent advances in large language models (LLMs) open up the possibility of generative CAD, where natural language is directly translated into parametric 3D models. However, most existing methods generate task-specific command sequences that pretrained models cannot directly handle. These sequences must be converted into CAD representations such as CAD vectors before a 3D model can be produced, which requires training models from scratch and adds unnecessary complexity. To tackle this issue, we propose generating CadQuery code directly from text, leveraging the strengths of pretrained LLMs to produce 3D models without intermediate representations, using this Python-based scripting language. Since LLMs already excel at Python generation and spatial reasoning, fine-tuning them on Text-to-CadQuery data proves highly effective. Given that these capabilities typically improve with scale, we hypothesize that larger models will perform better after fine-tuning. To enable this, we augment the Text2CAD dataset with 170,000 CadQuery annotations. We fine-tune six open-source LLMs of varying sizes and observe consistent improvements. Our best model achieves a top-1 exact match of 69.3%, up from 58.8%, and reduces Chamfer Distance by 48.6%. Project page: https://github.com/Text-to-CadQuery/Text-to-CadQuery.

Text-to-CadQuery: A New Paradigm for CAD Generation with Scalable Large Model Capabilities

TL;DR

This work targets the barrier to CAD generation by directly producing executable CadQuery code from natural language, bypassing intermediate command sequences. It introduces Text-to-CadQuery, a large-scale dataset of ~170k text–CadQuery pairs and a multi-model fine-tuning study that demonstrates a clear scaling trend: larger pretrained decoders yield better geometric fidelity and executability in CadQuery scripts. By leveraging a CadQuery-focused output and a data-annotated refinement loop, the approach achieves state-of-the-art results on standard CAD-geometric metrics, while enabling faster, code-based 3D generation without training from scratch. The findings suggest significant practical impact for scalable, language-driven CAD, with potential to extend to additional inputs and even larger models as data and compute resources grow.

Abstract

Computer-aided design (CAD) is fundamental to modern engineering and manufacturing, but creating CAD models still requires expert knowledge and specialized software. Recent advances in large language models (LLMs) open up the possibility of generative CAD, where natural language is directly translated into parametric 3D models. However, most existing methods generate task-specific command sequences that pretrained models cannot directly handle. These sequences must be converted into CAD representations such as CAD vectors before a 3D model can be produced, which requires training models from scratch and adds unnecessary complexity. To tackle this issue, we propose generating CadQuery code directly from text, leveraging the strengths of pretrained LLMs to produce 3D models without intermediate representations, using this Python-based scripting language. Since LLMs already excel at Python generation and spatial reasoning, fine-tuning them on Text-to-CadQuery data proves highly effective. Given that these capabilities typically improve with scale, we hypothesize that larger models will perform better after fine-tuning. To enable this, we augment the Text2CAD dataset with 170,000 CadQuery annotations. We fine-tune six open-source LLMs of varying sizes and observe consistent improvements. Our best model achieves a top-1 exact match of 69.3%, up from 58.8%, and reduces Chamfer Distance by 48.6%. Project page: https://github.com/Text-to-CadQuery/Text-to-CadQuery.
Paper Structure (17 sections, 4 equations, 4 figures, 4 tables)

This paper contains 17 sections, 4 equations, 4 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Method
  4. Data Annotation
  5. Pretrained Model Selection
  6. Fine-Tuning Setup
  7. Results and Evaluation
  8. Baseline: Text2CAD Transformer
  9. Evaluation Metrics
  10. Limitation
  11. Conclusion
  12. Technical Appendices and Supplementary Material
  13. Background: CAD Command Sequence Representation
  14. Example.
  15. Comparison and Examples of Parametric CAD Tools
  16. ...and 2 more sections

Figures (4)

  • Figure 1: Workflow comparison.
  • Figure 2: Overview of our data annotation pipeline.
  • Figure 3: Distribution of token lengths.
  • Figure 4: Comparison of F1 Score and IoU across models.