Answers to common questions about LensCAD.
On this page
What is LensCAD?
LensCAD is a parametric CAD platform for construction MEP and industrial design. You describe design intent in natural language or LCD code, and the system generates validated 2D/3D geometry, checks it against building codes and your own rules, and exports to eight industry formats.
What is LCD?
LCD (Lens CAD Description) is a domain-specific language that reads like a construction specification. It compiles to OpenCASCADE geometry operations. You can write it by hand or let the Agent generate it from a natural-language prompt.
How does the Agent work?
The Agent operates in two modes. In Plan Mode, it reads the brief, checks the knowledge base, and proposes a step-by-step layout plan for your review. In Edit Mode, it writes LCD code directly. Every change enters a review queue where you accept, reject, or revert line by line. The Agent suggests; you decide.
Is this another AI chatbot for construction?
No. LensCAD is a design execution system, not a Q&A assistant. The Agent writes executable CAD code that produces actual geometry, not text summaries. It references your project knowledge base and validates against building codes during generation.
What export formats are supported?
LensCAD exports to DXF, IGES, STEP, STL, GLTF, PDF, SVG, and 3MF. DXF and PDF for traditional CAD workflows, STEP and IGES for BIM/manufacturing, GLTF for web and AR, STL and 3MF for additive manufacturing.
What building codes are supported?
The platform includes reference tables and rule sets from NFPA 13, NFPA 72, IBC, ASHRAE 90.1, ASHRAE 62.1, NEC, IMC, and IPC. You can also upload your own code references and company standards into the per-project knowledge base.
How does the block library work?
Blocks are parametric components (sprinkler heads, diffusers, light fixtures, detectors, etc.) that carry their own rules: coverage area, clearance requirements, connection points, and circuit assignments. Place a block, and the rule engine validates spacing, clearance, and coverage automatically.
How does version control work?
Every save creates an immutable ProjectVersion snapshot. You can compare any two versions side by side, see semantic diffs (not just pixel diffs), and roll back without losing downstream work. This is design-intent versioning, not file-level versioning.
Can I import existing DWG/DXF files?
Yes. The Semantic CAD Parser reads DWG and DXF files, extracting layers, blocks, dimensions, and annotations. It understands what each entity represents and can convert existing drawings into LCD code for parametric editing.
What technology stack does LensCAD use?
The frontend runs on Next.js with TypeScript. The geometry kernel is OpenCASCADE compiled to WebAssembly, running directly in the browser with no server round-trip for geometry operations. 3D rendering uses Three.js. The backend uses PostgreSQL for project data and knowledge storage.
Can I self-host LensCAD?
Yes. Private deployment is available for enterprise teams that require on-premises infrastructure or data residency compliance. Contact us for details on self-hosted licensing.
Is my data private?
Your project data, knowledge bases, and generated designs are private to your account. We do not use customer data to train models. Private deployment customers have full data sovereignty with no external data transmission.
How do I get started?
Sign up and open the workspace directly from the website. No credit card required for the free tier. You can start designing immediately with the block library and LCD code editor.
What does pricing look like?
LensCAD offers a free tier for individual projects. Team and enterprise tiers include advanced knowledge base features, priority Agent capacity, private deployment options, and skill marketplace access. Contact us for enterprise pricing.