Blender Mcp Senpai
@xhiroga
Your AI-assisted mentor for Blender.
Overview
What is Blender MCP Senpai?
Blender MCP Senpai is an AI-assisted topology mentor for Blender that detects n-gons and topology issues in real time. It integrates with ChatGPT, Claude, or Gemini to provide improvement suggestions, and runs as a Blender add-on with a built-in MCP server (no external MCP process required). It is designed for 3D artists and modellers who want to improve their mesh topology.
How to use Blender MCP Senpai?
Install the add-on in Blender (requires Blender 4.2+). No further setup is needed for the basic topology detection. To enable AI suggestions, configure an MCP-compatible client (e.g., Claude Desktop or Dive) to connect to http://localhost:13180/sse using the JSON configuration provided in the README. After restarting the client, AI comments will appear for detected topology issues.
Key features of Blender MCP Senpai
- Auto n-gon highlight – instantly spots topology issues
- AI comments from ChatGPT, Claude, or Gemini
- Zero configuration – works out-of-the-box
- No external MCP server required
Use cases of Blender MCP Senpai
- Detecting n-gons and non-manifold geometry in 3D models
- Receiving AI-suggested improvements for mesh topology
- Learning best practices for clean topology during modelling
- Speeding up model review and iteration with real-time feedback
FAQ from Blender MCP Senpai
Do I need an external MCP server to run Blender MCP Senpai?
No. The add‑on includes its own MCP server component. You only need an MCP client (e.g., Claude Desktop) to connect to it.
Which AI models can provide suggestions?
ChatGPT, Claude, and Gemini are supported.
What version of Blender is required?
Blender 4.2 or later.
How do I configure an MCP client like Claude or Dive?
Add a JSON entry to your client as shown in the Quick Start section. For example, Claude uses npx -y supergateway --sse http://localhost:13180/sse; Dive uses SSE transport at the same URL. Restart the application after configuration.
What transport does the server use?
The add‑on communicates via SSE (Server‑Sent Events) on http://localhost:13180/sse.
