Connecting AI to Blender with the Model Context Protocol

Blender is an incredibly powerful 3D creation suite, but its Python API has a steep learning curve. What if you could just tell an AI what to build and have it control Blender directly?

That was the idea behind Blender MCP. It is an MCP server that acts as a bridge between AI assistants and Blender. Claude Code, GitHub Copilot, or any MCP-compatible client can issue commands that get executed in Blender in real-time.

The architecture is straightforward. The MCP server runs as a Node.js process and spawns Blender as a background subprocess. When a tool is invoked, the server generates a Python script using Blender's bpy API and executes it in the background Blender instance.

One of the interesting features is procedural model generation. The server includes templates for Indian-themed 3D models — temples with pillars and shikharas, auto-rickshaws with detailed wheels, traditional thalis with rice and curries. Each template is a parameterized Python script that generates geometry on the fly.

The biggest challenge was managing the Blender subprocess lifecycle. Blender can take seconds to start, and a long-running operation might timeout. I implemented a connection pool that keeps Blender running in the background and reconnects if it crashes.

Material generation was another interesting problem. Instead of using texture files, the server generates procedural materials using Blender's node system. Wood grain, marble, metal — each material is a node tree that gets created programmatically.

The project is open source on GitHub and published on npm as blender-mcp. It has been used to create everything from architectural visualizations to game assets.