IterCAD: An Iterative Multimodal Agent for Visually-Grounded CAD Generation and Editing
cs.AI, cs.CV
Submitted: 2026-06-11
Updated: 2026-08-31
Code: https://github.com/gumyr/build123d
License: http://creativecommons.org/licenses/by-nc-sa/4.0/
The gist: Computer-Aided Design is pivotal in modern manufacturing, yet existing automated methods predominantly rely on open-loop, one-shot generation, creating a mismatch with iterative real-world practices.
Terminology
Abstract
Computer-Aided Design is pivotal in modern manufacturing, yet existing automated methods predominantly rely on open-loop, one-shot generation, creating a mismatch with iterative real-world practices. In this paper, we present IterCAD, a unified multimodal agent framework for closed-loop, interactive CAD generation and editing. We formulate the task as a multi-turn interaction between a multimodal agent and an executable CAD sandbox, covering three tasks: Drawing-to-Code, Text-to-Code, and Interactive Editing. To support this, we develop a data synthesis pipeline incorporating advanced industrial manufacturing features to generate standard-compliant multi-view engineering drawings, complex code-editing tasks, and high-fidelity interaction trajectories. We optimize the agent via progressive SFT followed by geometry-aware reinforcement learning with viable-prefix masking to enhance code executability and geometric fidelity. Finally, we introduce the IterCAD-Bench evaluation suite and propose the Chamfer Distance Tolerance-Recall (CD-TR) curve alongside its AUC-TR metric, establishing a survivor-bias-free standard that unifies code validity and geometric precision. Extensive experiments demonstrate that IterCAD achieves highly competitive performance across multiple benchmarks, significantly outperforming existing approaches in both code executability and geometric precision, while exhibiting superior capabilities in closed-loop iterative refinement.
Sources
- Text-to-CadQuery: A New Paradigm for CAD Generation with Scalable Large Model Capabilities
- Clarify Before You Draw: Proactive Agents for Robust Text-to-CAD Generation
- SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebooks
- CADSmith: Multi-Agent CAD Generation with Programmatic Geometric Validation
- CME-CAD: Heterogeneous Collaborative Multi-Expert Reinforcement Learning for CAD Code Generation
- cadrille: Multi-modal CAD Reconstruction with Reinforcement Learning
- CADReasoner: Iterative Program Editing for CAD Reverse Engineering
- GIFT: Bootstrapping Image-to-CAD Program Synthesis via Geometric Feedback
- SolidGen: An Autoregressive Model for Direct B-rep Synthesis
- Seek-CAD: A Self-refined Generative Modeling for 3D Parametric CAD Using Local Inference via DeepSeek
- CAD-Editor: A Locate-then-Infill Framework with Automated Training Data Synthesis for Text-Based CAD Editing
- PR-CAD: Progressive Refinement for Unified Controllable and Faithful Text-to-CAD Generation with Large Language Models
- CADDesigner: Conceptual CAD Model Generation with a General-Purpose Agent
- TOOLCAD: Exploring Tool-Using Large Language Models in Text-to-CAD Generation with Reinforcement Learning
- Qwen3-VL Technical Report
- Group Sequence Policy Optimization
- SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement Learning
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems