LlamaCloud MCP Servers & Tools#
LlamaIndex provides official MCP servers that integrate with LlamaCloud services like LlamaCloud Indexes and LlamaExtract.
The llamacloud-mcp
Python package provides an alternative implementation that supports both query and extraction capabilities, using LlamaCloud indexes as knowledge bases and LlamaExtract agents for structured extraction.
Installation#
pip install llamacloud-mcp
*# or*
uvx llamacloud-mcp@latest
Claude Usage#
To use with an MCP host like claude-code, you can set your configuration file like so:
{
"mcpServers": {
"llama_index_docs_server": {
"command": "uvx",
"args": [
"llamacloud-mcp@latest",
"--index",
"your-index-name:Description of your index",
"--index",
"your-other-index-name:Description of your other index",
"--extract-agent",
"extract-agent-name:Description of your extract agent",
"--project-name",
"<Your LlamaCloud Project Name>",
"--org-id",
"<Your LlamaCloud Org ID>",
"--api-key",
"<Your LlamaCloud API Key>"
]
},
"filesystem": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-filesystem",
"<your directory you want filesystem tool to have access to>"
]
}
}
}
General Usage#
By default, the MCP server is launched with the stdio
transport. This is useful for hosts like Claude Desktop that support MCP servers via stdin/stdout, but you can also launch the server with the streamable-http
or sse
transport, which is useful for hosts that support MCP servers via HTTP.
You can launch the server directly from the command line:
llamacloud-mcp --index "index-name:Description" --extract-agent "name:description" --org-id YOUR_ORG_ID --project-id YOUR_PROJECT_ID --api-key YOUR_API_KEY --transport streamable-http