Blender AI MCP Server
Blender modeling operations executed via validated tool contracts instead of raw scripts.
Data last scanned yesterday · Reviewed last month
Overview
Blender AI MCP Server replaces unreliable raw Python script generation with a structured tool API for 3D modeling tasks. By utilizing predefined tools for inspection, measurement, and assertion, it ensures that AI clients interact with the Blender environment through stable, validated parameters rather than unpredictable code synthesis. The server features a goal-first orchestration pattern, allowing users to define high-level objectives before executing low-level operations. This approach limits the exposure of the entire Blender runtime to the AI, maintaining a smaller surface area for operations and utilizing internal logic to handle execution on the Blender main thread.
Our verdict
This server receives a security score of 30/100, resulting in a high-risk classification. The score reflects its nature as a community-maintained project rather than an official vendor release and the current absence of editorial security reviews. Potential users should be aware that the lack of verified authentication and read-only mode assessments necessitates cautious deployment in sensitive production environments.
- Uses stable tool contracts instead of generating raw bpy scripts.
- Implements goal-first routing for better task orchestration.
- Zero direct dependencies reduce the supply-chain surface area.
- Provides verification tools for inspecting the state of 3D scenes.
- Community-maintained project lacking official vendor oversight.
- Authentication methods have not undergone independent security review.
- Read-only mode implementation status remains unverified.
- Low bus factor with only a single contributor.
Tools
| Tool | Description | Risk |
|---|---|---|
| router_set_goal | Sets the active goal for the current session. | medium |
| router_get_status | Gets the status of the current goal and session. | low |
| browse_workflows | Browse available workflow tools. | low |
| search_tools | Search for available tools. | low |
| call_tool | Call a tool by name with arguments. | medium |
| check_scene | Queries the current scene state. | low |
| inspect_scene | Inspects a target object in the scene. | low |
| configure_scene | Configures scene settings. | medium |
| macro_cutout_recess | Performs recesses, openings, and cutter-driven cutouts. | medium |
| macro_relative_layout | Performs align/place/contact-gap part layout. | medium |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| ChatGPT | ||||
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf |
Frequently asked questions
›What are the environment requirements for running this server?
This server requires Python 3.11 or higher.
›Does this server support Docker?
Yes, the project provides a Docker-ready container.
›What is the recommended workflow for using the MCP server?
The recommended flow is to start with router_set_goal, followed by browsing workflows or searching tools, and finally verifying actions with inspection and measurement tools.
›How should I handle tool discovery during a session?
If a tool is not immediately visible, you should use the search_tools function before attempting to call the tool.
›Are there specific prompt assets I should use?
Yes, you should utilize the prompt-library assets located in the project's _docs/_PROMPTS/ directory as the canonical operating instructions.
Alternatives
Changelog
Badge
Maintain this server? Add the live badge to your README.