text-to-cad
Plugin that lets AI agents generate 3D CAD models (STEP, STL, GLB, 3MF) from text prompts.
- Stars
- 24.9k
- Forks
- 2.4k
- Licence
- MIT
- Language
- Python
- Last update
- 4 days ago
- Type
- AI model / agent
What it is
text-to-cad is a plugin for AI coding agents such as Claude Code, Codex, Cursor and Gemini. Your agent writes each 3D model as a Python script built on build123d and Open CASCADE, then runs it to produce files like STEP, GLB, STL or 3MF. It also includes a viewer that shows the model as it rebuilds, plus design-for-manufacturing checks and engineering drawings according to the README.
- Generate a STEP or STL file for a mechanical part from a text description
- Produce engineering drawings for a generated model
- Run design-for-manufacturing checks on a model
- Send a model to 3D printing, sheet metal or CNC fabrication services the README mentions
Great for
- Product designers who want to sketch physical parts by describing them in plain language
- Engineers exploring quick 3D model variations through an AI agent
- Hobbyists preparing STL or 3MF files for 3D printing
- Students learning parametric CAD through generated Python scripts
Not ideal for
- Users who need a traditional graphical CAD editor, since the README describes models being written as Python scripts by the agent
- Teams needing certified or production-verified engineering output, since the README does not address validation beyond basic checks
How to get it
- 1
Install with a package manager
You'll need: uv (Python package and tool manager) · Python 3.11+ runtime (launch commands use uv-managed Python 3.13) · A supported agent app such as Claude Code, Codex, Cursor, Grok Build or Gemini
Install in Claude Code (marketplace)
claude plugin marketplace add earthtojake/text-to-cad#latestInstall plugin in Claude Code
claude plugin install text-to-cad@earthtojakeInstall in Gemini
gemini extensions install https://github.com/earthtojake/text-to-cad --ref latest --consent --auto-updateClone for Cursor
git clone --depth 1 --branch latest https://github.com/earthtojake/text-to-cad ~/.cursor/plugins/local/text-to-cad - 2
Run it
- Confirm uv is installed by running uv --version
- Install the plugin for your agent app using the commands above
- Restart the agent app so it downloads the CAD runtime on first start
- Ask your agent to design a simple part and show it
- Open the CAD Viewer link if your app shows text instead of viewer cards
- 3
Get the most out of it
- Use the Claude Desktop config to pin a specific cadgen release for stable behaviour
- Turn telemetry off with the cadgen telemetry off command if you prefer no usage reporting
- Use the CAD tab in the Codex app to view models beside the thread
- Ask the agent to export STEP, GLB, STL or 3MF depending on your downstream tool
Downloads
Strengths
- Works with several popular agents that support plugins or the skills framework
- Exports common CAD and printing formats such as STEP, STL and 3MF
- Viewer updates as the model rebuilds
Trade-offs
- Output quality depends on the agent writing correct build123d code, which the README does not benchmark
- Requires uv and a network connection on first start to download the runtime
- Telemetry is on by default, though it can be turned off
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| text-to-cad (this) | 24.9k | AI model / agent | intermediate | 4 days ago | Product designers who want to sketch physical parts by describing them in plain language |
| langchain | 149k | Framework | intermediate | 4 days ago | Python developers building LLM-powered applications |
| crewAI | 60.4k | Library | intermediate | 4 days ago | Python developers prototyping multi-agent workflows |
| langgraph | 44.3k | Library | intermediate | 4 days ago | Python developers building multi-step AI agents that must survive failures or long waits |
| Prompt-Engineering-Guide | 79.8k | Learning resource | beginner | 8 months ago | People who are new to prompting and want structured lessons with examples |
| ai-engineering-from-scratch | 76.3k | Learning resource | intermediate | 4 days ago | Students who want to understand how LLMs, retrieval, and agents work internally |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.