Understand
How to Use Claude Code and the Official Blender MCP Server on a Cloud Computer
Overview
This guide covers how to set up and use Anthropic's Claude with Blender's official Model Context Protocol (MCP) server using a cloud-based Windows computer (Vagon) equipped with a dedicated GPU. It demonstrates how to handle a messy imported 3D project, inspect data blocks, clean and organize geometry, set up studio lighting, render batch images, and build custom tools using AI assistance while keeping local hardware cool.
Key points
- Blender features an official AI connector and MCP server built by its developers and announced by Anthropic for Claude.
- Because Claude and the Blender add-on must communicate locally over the same machine with a capable GPU, using a cloud computer like Vagon allows users to run heavy workflows from any laptop.
- Users can install the Claude desktop app on the cloud machine, connect it to Blender, and prompt Claude to analyze, summarize, and clean up inherited blend files safely.
- Claude can write and execute Python scripts inside Blender to set up studio lighting, add curved backdrops, place cameras, and generate screenshots to visually verify its work.
- Cloud performance tiers (such as Spark, Flame, Blaze, or Lava) can be scaled up or down dynamically depending on whether the user is doing everyday scene work or heavy rendering.
- Claude excels at tasks expressible in Python like structure, materials, lighting, and batch rendering, but remains less suited for manual artistry such as sculpting or detailed vertex work.