# CADSmith: Multi-Agent CAD Generation with Programmatic Geometric Validation arXiv: https://arxiv.org/abs/2603.26512 ยท HTML: https://mlsanity.com/paper/2603.26512/ ## Key facts - Cited by 13 papers in the map (data through 14 Oct 2026). - Placed in Multimodal and embodied agents, one of 16 directions in LLM agents and tool use. - Builds on 20 of its 33 references that are in the map. - First posted on arXiv 27 Mar 2026. ## Placed in - LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/) ## Builds on (most cited first, 20 of 20) - [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020) - [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021) - [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023) - [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023) - [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022) - [A Comprehensive Overview of Large Language Models](https://mlsanity.com/paper/2307.06435/) (2023) - [Evaluating Text-to-Visual Generation with Image-to-Text Generation](https://mlsanity.com/paper/2404.01291/) (2024) - [DeepCAD: A Deep Generative Network for Computer-Aided Design Models](https://mlsanity.com/paper/2105.09492/) (2021) - [Automatically Correcting Large Language Models: Surveying the landscape of diverse self-correction strategies](https://mlsanity.com/paper/2308.03188/) (2023) - [GPT-4V(ision) is a Human-Aligned Evaluator for Text-to-3D Generation](https://mlsanity.com/paper/2401.04092/) (2024) - [GPT-4V(ision) as a Generalist Evaluator for Vision-Language Tasks](https://mlsanity.com/paper/2311.01361/) (2023) - [CAD-MLLM: Unifying Multimodality-Conditioned CAD Generation With MLLM](https://mlsanity.com/paper/2411.04954/) (2024) - [3DALL-E: Integrating Text-to-Image AI in 3D Design Workflows](https://mlsanity.com/paper/2210.11603/) (2022) - [Query2CAD: Generating CAD models using natural language queries](https://mlsanity.com/paper/2406.00144/) (2024) - [OpenECAD: An Efficient Visual Language Model for Editable 3D-CAD Design](https://mlsanity.com/paper/2406.09913/) (2024) - [Large Language Models for Manufacturing](https://mlsanity.com/paper/2410.21418/) (2024) - [Text2CAD: Generating Sequential CAD Models from Beginner-to-Expert Level Text Prompts](https://mlsanity.com/paper/2409.17106/) (2024) - [Geometric Deep Learning for Computer-Aided Design: A Survey](https://mlsanity.com/paper/2402.17695/) (2024) - [Conceptual Design Generation Using Large Language Models](https://mlsanity.com/paper/2306.01779/) (2023) - [Sketch2Prototype: Rapid Conceptual Design Exploration and Prototyping with Generative AI](https://mlsanity.com/paper/2405.12985/) (2024) ## Built on by (most cited first, 13 of 13) - [IterCAD: An Iterative Multimodal Agent for Visually-Grounded CAD Generation and Editing](https://mlsanity.com/paper/2606.13368/) (2026) - [P3D-Bench: Benchmarking MLLMs for Parametric 3D Generation and Structural Reasoning](https://mlsanity.com/paper/2606.11152/) (2026) - [Autonomous Research Agents: A Survey of AI Scientists and the Verification Gap](https://mlsanity.com/paper/2608.05179/) (2026) - [MUSE: Benchmarking Manufacturable, Functional, and Assemblable Text-to-CAD Generation](https://mlsanity.com/paper/2605.28579/) (2026) - [Procedura: Agentic 3D Modeling with Procedural Control](https://mlsanity.com/paper/2608.26238/) (2026) - [Self-Improving CAD Generation Agents with Finite Element Analysis as Feedback](https://mlsanity.com/paper/2605.17448/) (2026) - [Nova3D: Code-Native Generation of Programmable 3D Assets](https://mlsanity.com/paper/2607.22738/) (2026) - [CADForge: Agentic Single-View CAD Reconstruction with Explicit Geometry Reasoning](https://mlsanity.com/paper/2610.04262/) (2026) - [Three-Phase Evaluation of AI-Assisted Software Development Life Cycle](https://mlsanity.com/paper/2607.05125/) (2026) - [Test-Time Scaling for CAD Generation via Verifier-Free Consensus Selection](https://mlsanity.com/paper/2608.09706/) (2026) - [Vision2CAD: A Visual Agent Harness for Explicit Geometry Referencing and Localization in Parametric CAD Modeling](https://mlsanity.com/paper/2609.22688/) (2026) - [AgenticCADedit: A Stateful, Tool-Mediated Agentic Approach to Multimodal 3D CAD Editing](https://mlsanity.com/paper/2609.29621/) (2026) - [VIGOR: Verifiable Industrial Geometry Optimization and Repair in Scene Programs](https://mlsanity.com/paper/2610.13145/) (2026) ## Abstract (arXiv) Existing methods for text-to-CAD generation either operate in a single pass with no geometric verification or rely on lossy visual feedback that cannot resolve dimensional errors. We present CADSmith, a multi-agent pipeline that generates CadQuery code from natural language. It then undergoes an iterative refinement process through two nested correction loops: an inner loop that resolves execution errors and an outer loop grounded in programmatic geometric validation. The outer loop combines exact measurements from the OpenCASCADE kernel (bounding box dimensions, volume, solid validity) with holistic visual assessment from an independent vision-language model Judge. This provides both the numerical precision and the high-level shape awareness needed to converge on the correct geometry. The system uses retrieval-augmented generation over API documentation rather than fine-tuning, maintaining a current database as the underlying CAD library evolves. We evaluate on a custom benchmark of 100 prompts in three difficulty tiers (T1 through T3) with three ablation configurations. Against a zero-shot baseline, CADSmith achieves a 100% execution rate (up from 95%), improves the median F1 score from 0.9707 to 0.9846, the median IoU from 0.8085 to 0.9629, and reduces the mean Chamfer Distance from 28.37 to 0.74, demonstrating that closed-loop refinement with programmatic geometric feedback substantially improves the quality and reliability of LLM-generated CAD models.