Unstructured API FAQs

Question 1

What data sources and destinations are supported?

Accepted Answer

The server supports various source connectors including S3, Google Drive, and Azure. Destination connectors include S3, Weaviate, Astra, Neo4j, and MongoDB, with more to come!

Question 2

Can I use Firecrawl with the MCP Server?

Accepted Answer

Yes! The server integrates with Firecrawl for web crawling and LLM text generation. You can use Firecrawl to retrieve HTML content or generate optimized text for large language models.

Question 3

What is the Unstructured API MCP Server?

Accepted Answer

It's an implementation that helps you interact with the Unstructured API, providing tools to manage your data sources, destinations, and workflows, such as creating connectors and running workflows.

Question 4

How do I install and run the MCP Server?

Accepted Answer

Install dependencies using `uv add "mcp[cli]"` and `uv pip install --upgrade unstructured-client python-dotenv`. Set your Unstructured API key in a `.env` file. Then, run the server using `mcp run uns_mcp/server.py` or `uv run uns_mcp/server.py`.

Question 5

How can I integrate the MCP server with Claude Desktop?

Accepted Answer

You can configure Claude Desktop to use the MCP server by creating a `claude_desktop_config.json` file in the Claude application support directory and specifying the command and arguments to run the server. Make sure your Unstructured API key is correctly configured.

Unstructured API

Unstructured API

주요 기능

사용 사례

주요 기능

사용 사례