CADAM is a text-to-CAD web application that generates 3D models from natural language descriptions and images using AI.
The tool addresses the friction of traditional CAD workflows by letting users describe what they want to build in plain English or by uploading reference images, then automatically generating parametric 3D models. It runs entirely in the browser using WebAssembly, eliminating installation overhead. The generated models are output as OpenSCAD code, which means they remain fully parametric and editable. Users can adjust dimensions through interactive sliders without requiring the AI to regenerate the entire model, and they can export results in multiple formats including STL, SCAD, and DXF. The application includes support for established OpenSCAD libraries like BOSL, BOSL2, and MCAD, and features real-time 3D preview powered by Three.js.
CADAM suits engineers, makers, and designers who want to rapidly prototype mechanical parts and assemblies without learning CAD software syntax. It works well for iterative design where you start with a natural language concept and refine dimensions through parametric controls. The project is positioned as an open-source alternative to proprietary text-to-CAD tools, though the README does not name specific competitors for direct comparison.
The project shows active development with a focus on expanding benchmark demonstrations of complex machines and multi-part assemblies generated from single prompts. Work spans both the AI generation pipeline and the browser-based interface, with attention to rendering performance and export reliability across multiple file formats.
