gNucleus Text-To-CAD MCP Server

Overview

The gNucleus Text‑To‑CAD MCP server bridges the gap between natural language design intent and parametric CAD models. By exposing a simple tool, it allows AI assistants to translate descriptive prompts—such as “draw a block with length=80 mm, width=40 mm, height=20 mm”—into fully‑editable CAD geometry. This capability eliminates the need for manual sketching or parametric file manipulation, making rapid prototyping and iterative design more accessible to engineers, designers, and even non‑technical stakeholders.

Problem Solved

Traditional CAD workflows require explicit modeling steps—defining dimensions, constraints, and features—often demanding specialized software knowledge. For teams working with large language models (LLMs) that support the Model Context Protocol, this creates a friction point: the assistant can generate textual specifications but cannot directly produce CAD files. The gNucleus server resolves this by acting as a mediator that converts plain text into structured CAD outputs, thereby enabling end‑to‑end automation from idea to design artifact.

Core Value for Developers

Developers integrating AI assistants into engineering pipelines benefit from a declarative interface: they can invoke the tool with minimal boilerplate, while the server handles authentication, request routing, and response formatting. The server’s API key management is straightforward—environment variables expose the host, key, and optional organization ID—allowing seamless deployment in CI/CD workflows or local development setups. By returning not only the CAD file but also a shared 3D viewer URL that expires after 24 hours, it provides an immediate visual feedback loop for reviewers.

Key Features

• Natural‑Language to CAD: Supports a wide range of prompts, from simple blocks to complex assemblies like differential gearboxes or tapered roller bearings.
• Multilingual Input: Accepts prompts in multiple languages (English, Chinese, Spanish, etc.), broadening accessibility for global teams.
• Parameter Extraction: Parses key design parameters (e.g., , ) and embeds them in the resulting model metadata.
• Secure Sharing: Generates time‑limited URLs for 3D visualization, facilitating quick stakeholder review without exposing sensitive data.
• Enterprise‑Ready: Optional organization ID support allows teams to manage API usage and billing centrally.

Use Cases

• Rapid Prototyping: Engineers can describe a component and instantly obtain a CAD file for 3D printing or simulation.
• Design Ideation: Product managers and designers can experiment with variations by tweaking textual prompts, accelerating concept iteration.
• Educational Tools: Instructors can demonstrate CAD concepts by asking the assistant to generate models on demand, making learning interactive.
• Documentation Automation: Technical writers can embed CAD generation steps directly into product manuals or design specifications.

Integration with AI Workflows

The MCP server is designed to fit seamlessly into existing LLM‑driven workflows. A Claude or other AI assistant can invoke as part of a larger conversation, passing the prompt and receiving a structured response that includes both the CAD file path and a viewer link. The server’s lightweight interface means it can be hosted locally, on a private cloud, or as part of an orchestrated micro‑service architecture. By exposing the tool through standard MCP tooling (e.g., Claude Desktop configuration), developers can add it to their assistant’s repertoire with a single JSON update, enabling instant access to generative CAD capabilities.

In summary, the gNucleus Text‑To‑CAD MCP server empowers developers to transform spoken or typed design intent into tangible CAD artifacts, streamlining the engineering workflow and unlocking new possibilities for AI‑assisted product development.