AI & text-to-CAD
MCP
Model Context Protocol
Definition
MCP is an open-source protocol for connecting AI applications to external tools, data, and workflows. In CAD it is used to let an LLM call a CAD program such as OpenSCAD, FreeCAD, or Fusion, run modeling commands, and read back results instead of generating geometry blind.
Also known as: Model Context Protocol, MCP server
Updated
How the protocol is structured#
The MCP introduction describes it as an open-source standard for connecting AI applications like Claude or ChatGPT to data sources, tools, and workflows. The specification defines three roles that exchange JSON-RPC 2.0 messages:
- Hosts: the LLM applications that start connections
- Clients: connectors inside the host application
- Servers: services that provide context and capabilities, such as tools the model can call
What an MCP server does for CAD#
A CAD MCP server wraps a CAD program's scripting interface as a set of tools. The model might call "create sketch", "extrude", "render preview", or "export STEP", and receive the result of each call before choosing the next one.
That feedback loop is the main difference from asking an LLM to write a whole CAD script at once. With an OpenSCAD MCP server, for example, the model can render a preview, check it, and revise the code. With a Fusion bridge, it can run API commands one at a time and see errors as they happen.
Vendor support#
Community MCP servers exist for OpenSCAD, FreeCAD, Blender, and Fusion. Vendors are starting to plan for it too: Autodesk's Fusion roadmap for 2026 says it will keep enabling extensibility through MCP tools and APIs for agent-driven workflows around Fusion Data.
Limits#
MCP only defines how the model and the tool talk to each other. The quality of the result still depends on the model's grasp of the CAD API and geometry, and on what the server exposes. Tools that can modify files or run code also need the same care as any other automation with write access to design data.
Sources
Related terms
LLM
An LLM is a large neural network trained on text to predict and generate language. In CAD, LLMs are used to write modeling code, drive a CAD program's API, answer software questions, or translate a prompt into structured modeling steps.
OpenSCAD
OpenSCAD is free, open-source CAD software where models are written as code using constructive solid geometry and extrusion of 2D outlines. Because the whole model is text, it is one of the easiest CAD targets for LLMs to generate and edit.
AI copilot
An AI copilot is an assistant built into CAD software that answers questions, suggests commands or best practices, and sometimes runs small tasks from a chat panel. Copilots help with the workflow around a model, and most do not generate a finished part from scratch.
Text-to-CAD
Text-to-CAD is AI that turns a natural-language description into editable CAD geometry, such as a STEP file, a parametric model, or CAD code. The goal is a model with exact faces and dimensions that can be modified and manufactured, not just a shape that looks right on screen.