Get started with the Unstructured Transform MCP server

Learn how to install the Unstructured Transform MCP server into various AI tools. You can then point to your files and have Transform start producing partitioned, enriched, chunked, and embedded data based on your files in minutes.

Tools and frameworks

The Unstructured Transform MCP server works with any AI tool that supports the Model Context Protocol (MCP) standard. Installation instructions are provided for the following popular AI tools:

• Claude Code
• Claude Desktop
• Cline
• Codex CLI
• Codex Desktop
• Cursor CLI
• Cursor IDE
• Devin CLI
• Devin Desktop
• DSPy
• Dust
• Firebase Genkit
• Google ADK (Agent Development Kit)
• Google Antigravity CLI
• Google Antigravity Desktop
• Google Gemini CLI
• Gumloop
• Goose
• IBM Bob IDE
• IBM Bob Shell
• LangChain / LangGraph
• Mastra
• Microsoft Agent Framework
• NVIDIA NeMo Agent Toolkit
• Pydantic AI
• Vercel AI SDK
• Visual Studio Code

The Transform MCP server is also listed in a variety of MCP registries, where supported MCP clients can discover and install it directly.

Other tools

For other AI tools that support the MCP standard, use the following command to connect:

npx -y mcp-remote https://mcp.transform.unstructured.io

This command might require Node.js to be installed on your local development machine. To see if it is, in your terminal, run node -v or node --version. Install Node.js.

To learn how to apply this command to a specific AI tool, and to determine whether Node.js must be installed on your local development machine, see your AI tool’s documentation.