6 MCP Servers for Using AI to Generate 3D Models
Written by
0 mins read
The Model Context Protocol (MCP) has revolutionized how AI assistants interact with external applications, moving beyond mere conversation to enable direct control of specialized software. For 3D modeling enthusiasts, architects, engineers, and designers, this technological leap means AI can now manipulate complex 3D modeling software through natural language commands. It can even support BIM modelling!
Rather than laboriously clicking through menus or learning specialized commands, you can simply tell Claude, "Create a cylindrical tower with a domed roof and place it at these coordinates," and watch as your favorite 3D modeling software executes the command.
I've identified six MCP servers specifically designed for 3D modeling applications, spanning everything from professional architectural tools to open source modelers and 3D printing platforms.
Professional CAD and BIM modeling tools
1. Revit MCP Server by revit-mcp (74 ⭐ on Github)
Revit’s MCP server connects AI assistants to Autodesk Revit, the industry-standard Building Information Modeling (BIM) software. It enables natural language control over complex architectural and engineering models.
Tools available
Tool name | Description |
| Fetches information about the currently active view in Revit. |
| Lists all elements visible in the current Revit view. |
| Retrieves available types of families (reusable components) in the project. |
| Gets details about elements currently selected by the user. |
| Creates elements placed by a single point (doors, windows, furniture). |
| Creates elements defined by a line (walls, beams, pipes). |
| Creates elements based on a surface (floors, ceilings). |
| Removes specified elements from the Revit model. |
| Potentially resets or cleans up parts of the model. |
| Changes properties of an existing element. |
| Searches for available custom modules or commands within the plugin. |
| Executes a specific pre-defined module or command. |
| Sends C# or Python code to be executed directly within Revit. |
| Applies colors to elements based on parameter values. |
| Automatically adds annotation tags to all walls in the current view. |
| Filters elements based on AI-provided criteria. |
| A generic tool to perform operations on elements. |
External APIs and technologies
Node.js: Runtime environment for the server component
TypeScript: For type-safe code structure
WebSockets (
wslibrary): Communication channel to the companion pluginJSON-RPC: Protocol layer over WebSockets
Zod: Schema definition and validation for data safety
Configuration requirements
No external API keys required
Required companion software: revit-mcp-plugin must be installed and running within Revit
Server expects the plugin to be accessible via WebSockets (default: localhost:8080)
Revit MCP excels at empowering AI to understand and manipulate complex building information models, making it invaluable for architectural workflows. The server works in tandem with its companion plugin, facilitating connection between AI assistants and Revit's powerful BIM environment.
2. AutoCAD LT AutoLISP MCP Server by hvkshetry (28 ⭐ on Github)
This ingenious server enables AI assistants to command AutoCAD LT (versions 2024+) by translating natural language into AutoLISP instructions, effectively adding automation capabilities to the lighter version of this venerable CAD platform.
Tools available
Tool name | Description |
| Checks connection to AutoCAD LT and initializes if needed. |
| Draws a straight line between two specified (X, Y) points. |
| Creates a circle based on a center point and radius. |
| Adds single-line text at a specified position. |
| Inserts a predefined block with options for scale, rotation, and ID. |
| Draws a line connecting two specified blocks using connection points. |
| Adds a text label near a specified block. |
| Places multiple blocks in sequence with defined spacing. |
| Creates a connected sequence of line segments from a point list. |
| Rotates a specific block around a given base point. |
| Adds a linear dimension annotation between points. |
| Applies a hatch pattern to a closed polyline. |
| Creates or modifies a layer with specified properties. |
| Executes raw AutoLISP code for advanced operations. |
External APIs and technologies
Python: Core server implementation
FastMCP: Framework for the MCP server
AutoLISP: Native scripting language for AutoCAD
pywin32: Interface with Windows API for AutoCAD window management
keyboard: Simulates keyboard input for command entry
pyperclip: Clipboard interaction for command transfer
Configuration requirements
No external API keys required
AutoCAD LT 2024 or newer: Earlier versions lack AutoLISP support
Window management considerations: AutoCAD window must remain visible and contain "AutoCAD LT" and "Drawing" in the title
Administrator privileges: May be required for proper window focus control
The AutoCAD LT MCP Server demonstrates impressive technical finesse in its approach to automation. By harnessing AutoLISP—newly available in AutoCAD LT 2024—it orchestrates complex operations through a combination of window management and command simulation techniques. The server's implementation navigates around AutoCAD LT's limitations by essentially "typing" commands on behalf of the user through the pywin32 and keyboard libraries.
General-purpose 3D modeling tools
3. BlenderMCP by ahujasid (10,475 ⭐ on Github)
This extraordinarily popular MCP server connects Blender—the free, open source 3D creation suite—with AI assistants, enabling natural language control over modeling, texturing, and even AI-powered 3D asset generation.
Tools available
Tool name | Description |
| Fetches detailed information about the current Blender scene. |
| Retrieves detailed information about a specific scene object. |
| Executes arbitrary Python code within Blender's environment. |
| Lists available asset categories from Poly Haven. |
| Searches for assets on Poly Haven with filtering options. |
| Downloads and imports a Poly Haven asset at chosen resolution. |
| Applies a previously downloaded Polyhaven texture to an object. |
| Checks if the Poly Haven integration is enabled. |
| Checks if the Hyper3D Rodin integration is enabled and reports its mode. |
| Generates a 3D model using Hyper3D Rodin based on text description. |
| Generates a 3D model using Hyper3D Rodin based on input images. |
| Checks the status of an ongoing Hyper3D Rodin generation task. |
| Imports a Rodin-generated 3D model into the Blender scene. |
| Helps define the preferred strategy for creating assets. |
External APIs and technologies
Python: For both the MCP server and Blender addon
FastMCP: Python library for MCP server development (
mcp[cli]>=1.3.0)Poly Haven API: Access to high-quality textures, HDRIs, and models
Hyper3D Rodin API: AI-powered 3D model generation from text/images
uv: Fast Python package installer/resolver
Requests: HTTP client for API interactions
Sockets (TCP): Communication between external server and Blender addon
Configuration requirements
Hyper3D Rodin API Key: Default free trial key included, but has daily limits
Options for API access:
Obtain the key from
hyper3d.aifor MAIN_SITE modeObtain credentials from
fal.aifor FAL_AI mode
Required software: Blender (>=3.0), Python (>=3.10)
Two-part setup:
MCP server configuration (external process)
Blender addon installation (within Blender)
BlenderMCP stands out for its extraordinary popularity and comprehensive approach. It combines traditional modeling capabilities with cutting-edge AI-generated asset creation through Hyper3D Rodin integration. The server's dual integration with Poly Haven for high-quality conventional assets and Rodin for AI-generated models creates a powerful ecosystem for both traditional and AI-assisted 3D creation.
4. SketchUp MCP by mhyrr (76 ⭐ on Github)
This clever integration connects the user-friendly SketchUp 3D modeling software to AI assistants, enabling natural language control over everything from basic geometry to complex woodworking joints.
Tools available
Tool name | Description |
| Create a new component (cube, cylinder, etc.). |
| Delete a component by its ID. |
| Move, rotate, or scale a component. |
| Get information about currently selected components. |
| Apply materials to a component. |
| Export the scene to various formats (SKP, OBJ, DAE, STL, images). |
| Execute arbitrary Ruby code within SketchUp. |
| Perform union, difference, or intersection operations. |
| Create beveled edges on a component. |
| Create rounded edges on a component. |
| Create a mortise and tenon joint between components. |
| Create a dovetail joint between components. |
| Create a finger joint (box joint) between components. |
External APIs and technologies
SketchUp Ruby API: Primary interface for SketchUp interaction
Python: For the MCP server implementation
Ruby: For the SketchUp extension component
TCP Sockets: Communication between the Python server and the Ruby extension
JSON-RPC: Protocol for structured message exchange
FastMCP: Framework for building MCP servers
uv: Python packaging tool mentioned for installation
Configuration requirements
No external API keys required
Two-part setup:
SketchUp extension installation (.rbz file)
MCP server configuration (for Claude or other clients)
Default communication: localhost:9876
SketchUp MCP distinguishes itself with its impressive range of woodworking-specific tools, making it particularly valuable for furniture design and crafting. The implementation of advanced joinery tools (mortise and tenon, dovetail, finger joint) demonstrates deep domain knowledge in both woodworking and 3D modeling. The server's architecture, with Python-to-Ruby communication via TCP sockets, elegantly navigates the complexity of cross-language interaction.
5. RhinoMcpServer by always-tinkering (31 ⭐ on Github)
This specialized server connects Rhino 3D—renowned for its NURBS modeling precision—to AI assistants, enabling natural language control over its powerful modeling capabilities.
Tools available
Tool name | Description |
geometry_tools.create_sphere | Creates a sphere with the specified center and radius. |
geometry_tools.create_box | Creates a box with specified dimensions and corner location. |
geometry_tools.create_cylinder | Creates a cylinder with specified base, height, and radius. |
scene_tools.get_scene_info | Retrieves information about scene objects. |
scene_tools.clear_scene | Deletes objects from the scene (all or current layer). |
scene_tools.create_layer | Creates a new layer within the Rhino document. |
log_claude_message | Logs a message from the AI assistant for debugging. |
External APIs and technologies
RhinoCommon API: Core .NET API for Rhino 3D interaction
.NET / C#: For the Rhino plugin component
Python: For the MCP server components
NLog: Logging framework for the C# plugin
FastMCP: Framework for the Python MCP server
JSON: Data format for inter-process communication
TCP Sockets: Communication channel between components
Configuration requirements
No external API keys required
Rhino Installation: Rhino 3D must be installed and running
Multiple server options:
Combined MCP Server (recommended): Single Python process
Socket-based Servers: Daemon + proxy architecture
Standalone Server: Older implementation
Extensive logging configuration: Structured logs in dedicated directories
The RhinoMcpServer project showcases an intricate architecture characteristic of complex CAD integrations. While currently under active development (with a note that it was facing issues with object creation), its combination of C# and Python components demonstrates a sophisticated approach to bridging the gap between Rhino's .NET environment and the Python-centric MCP ecosystem.
3D printing and model repositories
6. Thingiverse MCP Server by gpaul-mcp (0 ⭐ on Github)
This innovative server connects AI assistants directly to Thingiverse—the popular 3D printing model repository—enabling natural language search and discovery of printable 3D models.
Tools available
Tool name | Description |
| Searches for 3D models based on a search term. |
| Retrieves random 3D models from Thingiverse. |
| Fetches all available categories from Thingiverse. |
| Gets random 3D models from a specific category. |
| Gets detailed information about a specific model. |
| Searches for things with pagination and sorting options. |
| Gets featured, popular, newest, or staff-picked things. |
External APIs and technologies
Thingiverse API: Source of the 3D model data
Puppeteer: Node.js library for headless Chrome browser control
Puppeteer Extra + Stealth Plugin: For avoiding detection as automated scraping
Node.js: Runtime environment for the server
TypeScript: Type-safe JavaScript for robust code
Zod: Schema validation for parameter safety
Configuration requirements
Thingiverse App Token: Required for authentication
Must be obtained from the Thingiverse Developer Portal
Configured via
.env.developmentand.env.productionfiles
Build process notes: Special token handling during build
While the newest addition to the ecosystem (with 0 stars at the time of writing), the Thingiverse MCP Server demonstrates remarkable technical acumen in its approach to web integration. The server's use of Puppeteer with stealth plugins showcases an inventive approach to API access, likely circumventing rate limitations or authentication complexities. Its comprehensive toolset for model discovery and filtering demonstrates thoughtful consideration of how users might want to explore 3D printing resources through natural language.
Comparative analysis
Primary focus and requirements
MCP Server | Primary focus | Key technologies | API Key requirements | Best use cases |
Revit MCP | BIM/Building Design | Node.js, WebSockets, JSON-RPC | None, companion plugin required | Architectural modeling, building information management |
AutoCAD MCP | 2D/3D CAD | Python, AutoLISP, pywin32 | None, AutoCAD LT 2024+ required | Technical drawing, engineering diagrams, basic 3D modeling |
BlenderMCP | General 3D Creation | Python, Poly Haven API, Hyper3D | Optional Hyper3D API key | Character modeling, animation, rendering, and AI-generated assets |
SketchUp MCP | User-friendly 3D | Ruby, Python, JSON-RPC | None | Architectural visualization, woodworking, furniture design |
RhinoMcpServer | NURBS/Precision 3D | .NET/C#, Python, RhinoCommon | None, Rhino required | Industrial design, free-form surfaces, precision modeling |
Thingiverse MCP | 3D Printing Models | Node.js, Puppeteer, TypeScript | Thingiverse App Token | 3D print model discovery, printer project research |
Tool capabilities by category
MCP server | Tool count | Basic geometry | Material/texture | Advanced modeling | API access | Asset generation/access | Export/documentation |
Revit MCP | 17 | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ |
AutoCAD MCP | 14 | ✅ | ✅ | ✅ | ✅ | ❌ | ❌ |
BlenderMCP | 14 | ✅* | ✅ | ✅* | ✅ | ✅ | ✅* |
SketchUp MCP | 13 | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ |
RhinoMcpServer | 7 | ✅ | ❌ | ❌ | ✅ | ❌ | ❌ |
Thingiverse MCP | 7 | ❌ | ❌ | ❌ | ✅ | ✅ | ❌ |
* Through execute_blender_code or integrated services
MCP ecosystem integration support
If a compatibility is crossed out, it may still work – it was just not an explicitly mentioned compatibility in the repo.
MCP server | Claude Desktop | Cursor | Cline | Mentioned integrations |
Revit MCP | ✅ | ✅ | ❓ | Any MCP client |
AutoCAD MCP | ✅ | ✅ | ❓ | Potentially requires admin privileges |
BlenderMCP | ✅ | ✅ | ❓ | Requires Blender addon |
SketchUp MCP | ✅ | ✅ | ❓ | Requires SketchUp extension |
RhinoMcpServer | ✅ | ✅ | ❓ | Multiple server configurations |
Thingiverse MCP | ✅ | ✅ | ❓ | Network or direct command options |
Using these MCP servers together
The true potency of these MCP servers emerges when strategically combined, creating workflows that leverage each tool's strengths. Let's explore some compelling scenarios:
1. Architectural design workflow
Scenario: An architect developing a comprehensive building design from concept to presentation.
Begin with SketchUp MCP for rapid conceptual modeling and massing studies
Transition to Revit MCP for detailed building information modeling
Use BlenderMCP to apply photorealistic materials and lighting for visualization
Import specialized components from Thingiverse MCP for furniture and fixtures
Export final presentations from each platform as needed
This integrated workflow leverages SketchUp's conceptual speed, Revit's technical precision, Blender's visualization capabilities, and Thingiverse's vast component library—all accessible through natural language commands to your AI assistant.
2. Product design and prototyping with complex shapes
Scenario: A product designer developing a new consumer device from concept to prototype.
Start with RhinoMcpServer for precision NURBS modeling of the product form
Use BlenderMCP with Hyper3D Rodin to generate AI-assisted concept variations
Refine technical specifications with AutoCAD MCP
Source components or validate printability via Thingiverse MCP
Export for manufacturing or 3D printing from appropriate platforms
This workflow demonstrates how different specialized tools can contribute to a comprehensive product development process, from conceptual design through technical specification to manufacturing preparation.
3. Educational 3D modeling course
Scenario: An educator teaching 3D modeling concepts to students with various software preferences.
Use BlenderMCP for fundamental concepts (free, open source)
Demonstrate professional architectural workflows with Revit MCP
Showcase precision industrial design with RhinoMcpServer
Illustrate technical documentation with AutoCAD MCP
Explore real-world applications with Thingiverse MCP
This educational approach allows an AI assistant to demonstrate equivalent operations across multiple platforms, highlighting the underlying principles while accommodating different software preferences or career focuses.
The Future of AI-assisted 3D modeling
As the MCP ecosystem expands, we can anticipate even more sophisticated integrations spanning the full spectrum of 3D modeling applications. Whether you're designing buildings, products, characters, or mechanical components, these MCP servers offer a glimpse into a future where your tools adapt to your natural workflow rather than forcing you to adapt to theirs.
The revolution in 3D modeling isn't just about more powerful software—it's about more intuitive interfaces between human creativity and digital tools. MCP servers are a crucial step toward realizing that vision, making the power of 3D modeling more accessible, efficient, and collaborative than ever before.
If you’re interested in using GenAI to generate 3d models, but don’t want to use MCP servers, you might also check out projects like ComfyUI-3D-Pack. In this case, the models used would output a 3D model directly, rather than calling tools to construct models step by step. However, assets like these still run into issues with high poly counts, among other things.
How Snyk can help you secure your AI-generated 3D modeling workflows
AI tools bring exciting possibilities to 3D modeling, but security considerations remain important, especially when implementing custom code or sharing models online. Snyk can scan for vulnerabilities in your projects, protecting both your development environment and your creative assets.
For enterprise-grade security without the enterprise-level costs, open source projects can apply for free access to Snyk's premium tools through the Secure Developer project. Check out some of the projects that have already joined the initiative!
Get started with Snyk AI code security tools for free
No credit card required.
Create an account using Bitbucket and more options
By using Snyk, you agree to abide by our policies, including our Terms of Service and Privacy Policy.