Blender MCP Server

Blender MCP Server

Enables AI agents to fully control Blender through 50+ tools for 3D modeling, animation, materials, and scene management via HTTP endpoints.

Category
Visit Server

README

Blender MCP Server

Model Context Protocol (MCP) server for complete Blender control via API

This addon transforms Blender into an MCP server, exposing 50+ powerful tools that can be orchestrated by AI agents through HTTP endpoints. Perfect for AI-driven 3D workflows, automation, and creative experimentation.

✨ Features

  • šŸš€ 50+ Tools - Complete control over Blender's features:

    • Object creation, manipulation, and transformation
    • Material and shader system management
    • Animation and keyframe control
    • Camera and lighting setup
    • Modifiers and constraints
    • Physics simulations (rigid body, cloth, fluid)
    • Geometry nodes and procedural generation
    • File import/export operations
    • Scene optimization and batch operations
  • šŸ”’ Thread-Safe Execution - Enterprise-grade queue system for safe concurrent operations

  • šŸ“¦ Auto-Install Dependencies - Automatically installs required packages on first run

  • šŸŽ® Simple UI Panel - Start/stop server with one click from Blender's N-panel

  • šŸ“Š Real-time Monitoring - Track operations, statistics, and server status

  • šŸ”§ Production-Ready - Comprehensive error handling, logging, and caching

šŸš€ Quick Start

Installation

  1. Download blender_mcp.py
  2. Open Blender
  3. Go to Edit → Preferences → Add-ons
  4. Click the dropdown arrow (āŒ„) next to the search bar
  5. Select Install from Disk...
  6. Choose the downloaded blender_mcp.py file
  7. Enable the addon by checking the box next to "MCP Complete Server for Blender"

Starting the Server

  1. Press N in the 3D Viewport to open the sidebar
  2. Navigate to the MCP Server tab
  3. Click Start Server
  4. Server will start on http://localhost:8000

The addon will automatically install required dependencies on first run (FastAPI, Uvicorn, Pydantic, etc.).

šŸ¤– Using with PolyMCP

This MCP server is designed to work seamlessly with PolyMCP - a powerful framework for orchestrating MCP servers with AI agents.

Example: AI-Controlled Blender

#!/usr/bin/env python3
import asyncio
from polymcp.polyagent import UnifiedPolyAgent, OllamaProvider

async def main():
    # Initialize your LLM provider
    llm = OllamaProvider(model="gpt-oss:120b-cloud", temperature=0.1)
    
    # Connect to Blender MCP server
    agent = UnifiedPolyAgent(
        llm_provider=llm, 
        mcp_servers=["http://localhost:8000/mcp"],  
        verbose=True
    )
    
    async with agent:
        print("āœ… Blender MCP Server connected!\n")
        
        # Chat with your AI to control Blender
        while True:
            user_input = input("\nšŸŽØ You: ")
            
            if user_input.lower() in ['exit', 'quit']:
                break
            
            result = await agent.run_async(user_input, max_steps=5)
            print(f"\nšŸ¤– Blender: {result}")

if __name__ == "__main__":
    asyncio.run(main())

Example Commands

Once connected, you can ask the AI agent to:

  • "Create a cube at position (0, 0, 0) with size 2"
  • "Add a red metallic material to the selected object"
  • "Create a camera looking at the origin"
  • "Set up a simple lighting scene with 3 lights"
  • "Add a fluid simulation to the cube"
  • "Export the scene as an FBX file"
  • "Create an animation rotating the object 360 degrees over 100 frames"

That's it! PolyMCP handles all the complexity of:

  • Tool discovery and selection
  • Multi-step task planning
  • Error handling and retries
  • State management across operations

This makes it incredibly simple to build AI-powered Blender automation tools!

šŸ“” API Endpoints

Once the server is running, you can access:

  • API Documentation: http://localhost:8000/docs
  • List All Tools: http://localhost:8000/mcp/list_tools
  • Invoke Tool: POST http://localhost:8000/mcp/invoke/{tool_name}

šŸ”§ Configuration

You can customize the server by editing the Config class in blender_mcp.py:

class Config:
    HOST = "0.0.0.0"           # Server host
    PORT = 8000                # Server port
    AUTO_INSTALL_PACKAGES = True  # Auto-install dependencies
    THREAD_SAFE_OPERATIONS = True  # Enable thread-safe execution
    ENABLE_CACHING = True      # Enable result caching

šŸ“‹ Requirements

The addon automatically installs these dependencies:

  • FastAPI
  • Uvicorn
  • Pydantic
  • docstring-parser
  • NumPy
  • PolyMCP

Blender Version: 3.0.0 or higher

šŸ› Troubleshooting

Server won't start?

  • Check Blender's System Console for error messages (Window → Toggle System Console)
  • Ensure port 8000 is not already in use
  • Try restarting Blender after installation

Dependencies not installing?

  • Manually install packages using Blender's Python:
    /path/to/blender/python -m pip install fastapi uvicorn pydantic docstring-parser numpy
    

Can't find MCP Server panel?

  • Press N in the 3D Viewport
  • Look for "MCP Server" tab in the sidebar
  • Make sure the addon is enabled in Preferences

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured