dcc-mcp-blender: AI-Driven 3D Workflows with an Embedded MCP Server

Introduction

dcc-mcp-blender transforms Blender into a first-class server within the DCC Model Context Protocol (MCP) ecosystem. This innovative addon embeds a Streamable HTTP MCP server directly inside Blender's Python interpreter, eliminating the need for external gateways. With dcc-mcp-blender, AI agents and other MCP clients can directly interact with and drive your Blender projects, enabling advanced automation and intelligent assistance for 3D content creation.

Why Use dcc-mcp-blender and Key Benefits

dcc-mcp-blender is designed for robust production pipelines and cross-DCC automation, offering significant advantages for developers and 3D artists leveraging AI. Unlike other solutions, its embedded server architecture ensures tight integration and efficiency. It provides over 200 pre-built, schema-validated tools across more than 25 skill packages, covering everything from scene management and object manipulation to rendering and physics. The extensible skill system allows for easy integration of new functionalities, and critical security features, such as the ability to disable arbitrary script execution, provide greater control over agent capabilities. This makes dcc-mcp-blender ideal for scenarios requiring precise, testable, and secure AI-driven workflows, including headless and CI/CD environments. Its compatibility with the shared MCP gateway and CLI also facilitates seamless integration with other DCC applications like Maya, Houdini, and 3ds Max.

Installation

Blender Compatibility: The addon targets Blender 5.2.1 LTS (Python 3.13) and Blender 4.5.13 LTS (Python 3.11), with legacy support for older versions. Ensure your Blender version is compatible.

Recommended: Install as Blender Extension (ZIP)

This is the only supported GUI installation path for Blender 4.2+.

1. Download the latest platform-specific ZIP from the dcc-mcp-blender Releases page. Look for dcc_mcp_blender_addon_win64_vX.Y.Z.zip, dcc_mcp_blender_addon_linux_vX.Y.Z.zip, or dcc_mcp_blender_addon_macos_vX.Y.Z.zip.
2. In Blender 4.2+, navigate to Edit ? Preferences ? Extensions ? Install from Disk… and select the downloaded ZIP file.
   • Important: Do NOT use Edit ? Preferences ? Add-ons ? Install, as this legacy path is unsupported for extension packages.
3. Enable the DCC MCP Blender extension.
4. The MCP server will automatically start on an OS-assigned instance port and register with the local gateway.

Alternative: Install via pip (for scripts / CI)

pip install dcc-mcp-blender

Then, in Blender's Python console:

import dcc_mcp_blender

# Start the server on an OS-assigned instance port
dcc_mcp_blender.start_server()

# Stop the server
dcc_mcp_blender.stop_server()

Examples

Using with Claude Desktop

To connect Claude Desktop, add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "blender": {
      "url": "http://127.0.0.1:9765/mcp"
    }
  }
}

Ensure the Blender addon is enabled and running, then restart Claude Desktop.

Agent Workflow with dcc-mcp-cli

AI agents can interact with Blender using the dcc-mcp-cli. First, ensure the CLI is installed and updated:

# Linux/macOS
curl -fsSL https://raw.githubusercontent.com/dcc-mcp/dcc-mcp-core/main/scripts/install-cli.sh | sh

# Windows PowerShell
powershell -ExecutionPolicy Bypass -c "irm https://raw.githubusercontent.com/dcc-mcp/dcc-mcp-core/main/scripts/install-cli.ps1 | iex"

dcc-mcp-cli update apply

Then, you can ask your agent to interact with Blender, for example:

Use dcc-mcp to inspect the current Blender scene.

Verify the Blender instance is registered:

dcc-mcp-cli list

Impressive Showcases

dcc-mcp-blender has been used to create stunning visual showcases, demonstrating its capabilities in complex 3D tasks. These include a weathered crate with detailed material reconstruction, stylized reference reconstruction with editable geometry, and rain-soaked PBR LookDev for a sports car, showcasing asset import, particle simulations, and advanced rendering through MCP.

Links

• GitHub Repository: dcc-mcp/dcc-mcp-blender
• Releases: Latest Releases
• PyPI: dcc-mcp-blender on PyPI
• DCC Model Context Protocol (MCP): dcc-mcp/dcc-mcp-core
• Website Gallery: DCC-MCP Showcase