Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
Blender MCP is an open-source MCP tool that connects large language models to Blender. Through environments such as Claude Desktop, Cursor, Windsurf, VSCode Roo Cline, and Trae Editor, it turns natural-language instructions into Blender actions or Python scripts. It can be used to create objects, modify materials, adjust lighting and cameras, take screenshots, call Polyhaven assets, and upload work to Sketchfab.
Its key value is not replacing Blender, but handing repetitive operations, script writing, and scene setup in the modeling workflow over to AI assistance. The text indicates that it supports not only Claude, but also models such as DeepSeek R1, Gemini, and Qwen via OpenRouter, as well as local models through Ollama. Claude is better suited to creative expression, while DeepSeek leans more toward technical tasks and code generation. For complex tasks, it can also be combined with server-sequential-thinking to break work down into steps.
Blender MCP itself is open-source and free, and can be used for both personal and commercial projects. However, if you use Claude, you will need Anthropic/Claude access or a subscription. When using services such as OpenRouter or Trae, costs depend on the respective platform. The page also explicitly mentions that you can use VSCode with free models on OpenRouter, or use local Ollama models to achieve zero model API-call costs.
Its main advantage is that it covers a fairly complete workflow: natural-language modeling, Python execution, material and scene control, asset integration, screenshots, and sharing are all included, and it is compatible with multiple editors. The downside is that the installation chain is relatively long, requiring Blender, Python, uv, plugins, and MCP client configuration. The quality of complex modeling results depends heavily on prompt design and model capability, and often requires multiple rounds of iteration. The text also notes that some viewport operations are limited; for example, object centering/focusing commands are not always reliable.
It is suitable for concept artists, game developers, architectural visualization professionals, educators, and technical creators who are willing to use AI to generate Blender Python scripts. For complete beginners, it is still recommended to first learn the basic concepts of Blender. The text does not specify access conditions from China; however, external services such as Claude, OpenRouter, Sketchfab, and GitHub may involve uncertainty around network access and payments. Local Ollama, domestic/open-source models, or traditional Blender plugins can be considered as alternative paths.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on blender-mcp.com official site.
blender-mcp.com is an Unknown AI Apps provider. TG4G tracks its product information, an overall rating of 7.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach blender-mcp.com directly.