Getting Started
Install the Blender MCP Server addon, start the HTTP server, and connect your AI client in minutes.
This guide walks you through installing the Blender MCP Server addon, verifying it is running, and connecting an AI client.
Prerequisites
- Blender 5.0 or later
- Windows, macOS, or Linux
- An MCP-compatible AI client (Claude Code, Cursor, or similar)
Step 1: Install the Addon
- Download the
blender_mcp_server.zipfile from your purchase. - Open Blender and go to Edit > Preferences > Add-ons.
- Click Install from Disk and select the downloaded zip file.
- Enable the addon by checking the checkbox next to Blender MCP Server.
For more details and alternative installation methods, see the Installation page.
Step 2: Verify the Server Is Running
The MCP server starts automatically when the addon is enabled. Check the Blender System Console (Window > Toggle System Console on Windows, or launch Blender from a terminal on macOS/Linux) for these messages:
BlenderMCP: === Blender MCP Server starting up ===
BlenderMCP: Registered 212 tools (preset: Full)
BlenderMCP: MCP server ready with 212 tools on port 13580
BlenderMCP: Connect your AI client to: http://localhost:13580/mcp
BlenderMCP: === Blender MCP Server startup complete ===
You can also verify the server status in the addon preferences panel under Edit > Preferences > Add-ons > Blender MCP Server, which displays the current port and tool count.
Step 3: Connect Your AI Client
Claude Code
Add the following to your MCP configuration file (.claude/mcp.json in your project directory):
{
"mcpServers": {
"blender": {
"type": "url",
"url": "http://localhost:13580/mcp"
}
}
}
Cursor
Add to Cursor's MCP settings (Settings > MCP):
{
"mcpServers": {
"blender": {
"url": "http://localhost:13580/mcp"
}
}
}
Other MCP Clients
Any MCP-compatible client can connect via Streamable HTTP:
- Endpoint:
http://localhost:13580/mcp - Method: POST (JSON-RPC 2.0 requests)
- Content-Type:
application/json
Legacy SSE transport is also supported:
- SSE endpoint:
GET http://localhost:13580/sse - Message endpoint:
POST http://localhost:13580/message
Step 4: Test the Connection
Once connected, ask your AI client to run a simple command such as:
"List all objects in the scene"
The client will call the get_scene_info tool and return the results. You can also try:
"Create a cube at position 0, 0, 2"
"Set the render resolution to 1920x1080"
What to Read Next
- Installation -- Configuration options, addon preferences, and troubleshooting
- Tool Categories -- Browse all 212 tools across 22 categories
- Presets -- Choose the right preset for your workflow
- Resources -- Read-only data sources for scene context
- Prompts -- Reusable workflow templates