ParseFlow

You can run and interact with the ParseFlow MCP Server to enable AI-powered document parsing across PDFs, Word, Excel, PowerPoint, and OCR images. This server exposes a set of tools you can call through an MCP client to extract, search, and semantically index content from diverse documents, making it easy to build intelligent assistants and workflows around your files.

How to use

Begin by starting the MCP server locally or remotely, then connect to it with an MCP client. You can execute a variety of document processing actions such as extracting text, converting formats, performing full-text searches, or generating semantic indices. The server supports multiple document types and can handle batch processing with parallel execution to speed up workflows. You will typically call individual tools to perform a specific task on a given file and then combine results within your application.

How to install

Prerequisites: ensure you have Node.js and npm installed on your machine.

# Install the core library for parsing and handling documents
npm install parseflow-core

# Install the MCP server globally for quick start
npm install -g parseflow-mcp-server

# Or run it directly with npx without installing globally
npx parseflow-mcp-server

Additional setup and configuration

To enable the MCP server within your environment, you can configure Claude Desktop to run the ParseFlow MCP server as part of its workflow. The following configuration snippet demonstrates how to add the server to Claude Desktop’s MCP configuration.

{
  "mcpServers": {
    "parseflow": {
      "command": "npx",
      "args": ["-y", "parseflow-mcp-server"]
    }
  }
}

Available tools

extract_text

Extracts plain text from documents or images and supports per-page operations for PDFs.

get_metadata

Retrieves metadata from documents, such as author, title, and creation date.

search_pdf

Performs full-text search within PDF documents.

extract_images

Extracts embedded images from documents.

get_toc

Retrieves the table of contents from PDF documents.

merge_pdf

Merges multiple PDFs into a single document.

split_pdf

Splits a PDF into single-page documents.

extract_pdf_pages

Extracts specific pages from a PDF by page range.

add_watermark

Adds a text watermark to PDF pages.

add_image_watermark

Adds an image watermark to PDF pages.

remove_watermark

Removes watermarks from PDF pages.

extract_word

Extracts text and optionally HTML from Word documents.

search_word

Searches within Word documents for specified terms.

extract_excel

Extracts data from Excel workbooks.

search_excel

Searches within Excel cells for specified terms.

extract_powerpoint

Extracts slides and text from PowerPoint presentations.

search_powerpoint

Searches across slides in a PowerPoint deck.

extract_ocr

Performs OCR on images to extract text.

search_ocr

Searches OCR-derived text for keywords.

semantic_index

Creates vector embeddings for documents to enable semantic search.

semantic_search

Performs semantic similarity search against indexed documents.

batch_extract

Performs extraction across multiple files in parallel.

batch_search

Performs search across multiple files in parallel.