Lune

CVPR Findings2026

CADReasoner: Iterative Program Editing for CAD Reverse Engineering

Soslan Kabisov, Vsevolod Kirichuk, Andrey Volkov, Marina Barannikov, Gennadiy Savrasov, Anton Konushin, Andrey Kuznetsov, Dmitrii Zhemchuzhnikov

Abstract

Computer-Aided Design (CAD) powers modern engineering, yet producing high-quality parts still demands substantial expert effort. Many AI systems tackle CAD reverse engineering, but most are single-pass and miss fine geometric details. In contrast, human engineers compare the input shape with the reconstruction and iteratively make changes in the design based on the remaining differences. Agentbased methods mimic this loop with frozen VLMs, but weak 3D grounding of current foundation models limits reliability and efficiency. We introduce CADReasoner, a model trained to iteratively refine its prediction based on the geometric discrepancy between the input and the predicted shape. The model outputs a runnable CadQuery Python program, feeding its rendered mesh back to the next step. CADReasoner fuses multi-view renders and point clouds as complementary modalities. To bridge the realism gap, we propose a scan-simulation protocol applied during both training and evaluation. Across DeepCAD, Fusion 360, and MCB benchmarks, CADReasoner attains state-of-theart results on clean and scan-sim tracks. Code and the model are available at GitHub and Hugging Face model.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Reflexion: language agents with verbal reinforcement learning cites CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning
  • Reflexion: language agents with verbal reinforcement learning cites Self-Refine: Iterative Refinement with Self-Feedback
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on35

  1. Reflexion: language agents with verbal reinforcement learningNoah Shinn, Federico Cassano, Ashwin Gopinath, Karthik Narasimhan et al.NeurIPS 2023 · 5,828 citations
  2. Self-Refine: Iterative Refinement with Self-FeedbackAman Madaan, Niket Tandon, Prakhar Gupta, Skyler Hallinan et al.NeurIPS 2023 · 4,972 citations
  3. Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?Zhiqi Chen, Rui Lu, Andrew Zhao, Zhaokai Wang et al.NeurIPS 2025 · 1,109 citations
  4. CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement LearningHung Le, Yue Wang, Akhilesh Deepak Gotmare, Silvio Savarese et al.NeurIPS 2022 · 571 citations
  5. DeepCAD: A Deep Generative Network for Computer-Aided Design ModelsRundi Wu, Chang Xiao, Changxi ZhengICCV 2021 · 290 citations

Related papers

  1. CAD-Refiner: A Unified Framework for CAD Generation and Iterative EditingMeng Yuan, Dawei Lin, Hongxia Xie, Tieru Wu et al.CVPR 2026
  2. ReACT: Reward-informed Autoregressive Decision CAD TransformerYijie Ding, Yang Liu, Haobo Jiang, Jianmin ZhengAAAI 2026 · 1 citation
  3. CAD-Recode: Reverse Engineering CAD Code From Point CloudsDanila Rukhovich, Elona Dupont, Dimitrios Mallis, Kseniya Cherenkova et al.ICCV 2025 · 16 citations
  4. MiCADangelo: Fine-Grained Reconstruction of Constrained CAD Models from 3D ScansAhmet Serdar Karadeniz, Dimitrios Mallis, Danila Rukhovich, Kseniya Cherenkova et al.NeurIPS 2025 · 4 citations
  5. CAD-GPT: Synthesising CAD Construction Sequence with Spatial Reasoning-Enhanced Multimodal LLMsSiyu Wang, Cailian Chen, Xinyi Le, Qimin Xu et al.AAAI 2025 · 49 citations
Dusk over the sea between two cliffs drawn in fine vertical lines