Blender MCP with Google Gemini — Setup Guide

The Google Gemini CLI (gemini-cli) supports MCP servers natively, which means you can connect Blender MCP to Gemini for AI-powered 3D modeling. This guide walks you through the complete setup — from installing the Gemini CLI to creating your first 3D scene with natural language.

The MCP configuration for Gemini CLI is similar to other clients: you add the Blender MCP server to a config file, and Gemini discovers and uses its tools automatically. Setup takes about 10 minutes.

Prerequisites

• Blender 3.0+ — Download from blender.org 
• Google Gemini CLI — Install via npm or your package manager
• uv package manager — Required for running the Blender MCP server
• Google AI API key — Free tier available at aistudio.google.com 

Step 1: Install Gemini CLI and uv

npm install -g @google/gemini-cli

curl -LsSf https://astral.sh/uv/install.sh | sh

npm install -g @google/gemini-cli

powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

$localBin = "$env:USERPROFILE\.local\bin"
$userPath = [Environment]::GetEnvironmentVariable("Path", "User")
[Environment]::SetEnvironmentVariable("Path", "$userPath;$localBin", "User")

Step 2: Configure Blender MCP in settings.json

The Gemini CLI reads MCP servers from ~/.gemini/settings.json. Create or edit this file and add the Blender MCP server to the mcpServers block:

{
  "mcpServers": {
    "blender": {
      "command": "uvx",
      "args": ["blender-mcp"]
    }
  }
}

{
  "mcpServers": {
    "blender": {
      "command": "cmd",
      "args": ["/c", "uvx", "blender-mcp"]
    }
  }
}

Step 3: Install the Blender Addon

The Blender addon creates a socket server inside Blender that receives commands from Gemini.

Step 4: Start the Blender Connection

Step 5: Verify and Use

Start the Gemini CLI:

gemini

Verify the Blender MCP server is connected:

/mcp list

You should see something like ✓ blender: uvx blender-mcp - Connected. If it shows Disconnected, run /mcp reload and ensure Blender is running with the addon connected.

Your first 3D prompt

Ask Gemini in natural language. For example:

• “Create a red metallic sphere above a blue cube in Blender”
• “Set up studio lighting and point the camera at the scene”
• “Create a low poly dungeon scene with a dragon”
• “Download a forest HDRI from Poly Haven and set it as the world background”

Gemini will use the Blender MCP tools to execute your request. You don’t need special syntax — just describe what you want.

Troubleshooting

”blender” shows Disconnected in /mcp list

• Make sure Blender is open, the addon is enabled, and you clicked Connect to Claude
• Verify uv is installed: uv --version
• On Windows, ensure the config uses the cmd + /c wrapper
• Try /mcp reload to force Gemini to re-connect

Gemini can’t find Blender tools

• Run /mcp reload to re-query the server for tools
• Restart the Gemini CLI after editing settings.json
• Confirm the BlenderMCP addon shows a connected status in Blender

First command times out or fails

This can happen with the first request after connecting. Send a second message — it usually works afterward.

API key / authentication

The Gemini CLI needs a Google AI API key for the model. Set it via gemini config set api_key YOUR_KEY or through the AI Studio  setup flow. Blender MCP itself requires no API key.

For a deep dive on how the server works, see our MCP Server architecture guide.