3D

This Skill addresses the fragmentation of cross-platform operations in 3D creation. In practical development, developers or content creators often need to switch between different 3D software such as Blender, Maya, and Unreal Engine, each with its own scripting API and operational methods, making it difficult for LLMs to invoke them uniformly. 3D-MCP provides a standardized TypeScript interface via the Model Context Protocol, allowing LLMs to operate multiple 3D platforms with the same set of instructions, eliminating the need for individual adaptation per software.

Usage is straightforward: simply install the corresponding plugin in the target 3D software (e.g., the Blender plugin) and connect via an MCP client (such as Claude or Cursor). You can then use natural language commands to create models, set animations, adjust materials, etc., for example, "Create a rotating cube in Blender and add keyframe animation." All operations are automatically mapped to the respective platform's API through a unified atomic action layer (e.g., createKeyframes).

It is well-suited for teams or individuals requiring cross-platform 3D workflows, particularly those already using multiple 3D software for content creation, game development, or film production, as well as developers looking to seamlessly integrate AI capabilities into their 3D pipeline. For agent developers, it is also an ideal tool, enabling AI agents to directly manipulate 3D scenes.

We recommend using this tool when you need to uniformly manage operations across multiple 3D software. Note that it primarily relies on plugin support for each platform; currently, Blender, Maya, and Unreal Engine are officially supported, while other platforms require custom adapter development. Additionally, complex high-level operations (such as physics simulations) may require combining multiple atomic actions to achieve.