# Papyrio PDF Tools [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/atabekdemurtaza/papyrio-mcp  
**GitHub Stars:** 1  
**npm Downloads (last month):** 1310  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/papyrio-pdf-tools

## Description
Compress, convert, merge, split and OCR PDFs from an assistant, using local files.

## Tools
Capabilities this server exposes over MCP:

- **compress_pdf** — .pdf
- **merge_pdfs** — .pdf
- **split_pdf** — .zip
- **extract_pdf_text** — .txt
- **pdf_to_word** — .docx
- **pdf_to_jpg** — .zip
- **ocr_pdf** — .pdf
- **office_to_pdf** — .pdf
- **image_to_pdf** — .pdf
- **webpage_to_pdf** — .pdf
- **protect_pdf** — .pdf
- **unlock_pdf** — .pdf

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "papyrio-pdf-tools": {
    "command": "npx",
    "args": ["-y","papyrio-mcp"],
    "env": {
      "PAPYRIO_API_KEY": ""
    }
  }
}
```

**Requires environment variables:** `PAPYRIO_API_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What Papyrio PDF Tools does

Papyrio PDF Tools exposes the Papyrio PDF API through the Model Context Protocol. It lets an MCP-compatible assistant work with files on the local machine using absolute paths. Operations include reducing PDF size, joining documents, separating pages, extracting existing text, converting between PDF and common document or image formats, adding OCR, rendering webpages, and applying or removing password protection.

The tool set supports workflows that combine several operations. For example, an assistant can OCR scanned invoices, merge the resulting PDFs, and compress the combined file as a sequence of tool calls. Output files are written locally, with most tools deriving a name from the input and adding a `-papyrio` suffix when `output_path` is omitted.

## How it works

The MCP process receives tool requests from the assistant and makes corresponding requests to the Papyrio PDF API. Input and output locations are local absolute paths. Although the workflow is controlled from the local machine, the files are uploaded to the Papyrio API for processing in memory. The API discards them after returning the result.

The server can run without an API key through a guest tier limited to three operations per day per IP address. A RapidAPI key enables the paid API path; the free RapidAPI tier allows 50 requests per month and does not require a payment card.

## Setup and configuration

For Claude Desktop, download the latest `papyrio-mcp.mcpb` release and open it to install the extension. The API key can be left empty for guest access or added later in the extension settings.

For a manual installation, configure the server with `npx`:

```json
{
  "mcpServers": {
    "papyrio": {
      "command": "npx",
      "args": ["-y", "papyrio-mcp"]
    }
  }
}
```

Set `PAPYRIO_API_KEY` in the server's `env` object when using a RapidAPI key. The package can also be built locally with `npm install` and `npm run build`; a client can then launch `dist/index.js` with Node.js.

## Tools and capabilities

- `compress_pdf` reduces PDF size, particularly for image-heavy documents.
- `merge_pdfs` joins between two and ten PDFs in the supplied order.
- `split_pdf` places each PDF page into a ZIP archive.
- `extract_pdf_text` returns the document's existing text layer as UTF-8.
- `pdf_to_word` creates an editable DOCX reconstruction.
- `pdf_to_jpg` renders PDF pages into a ZIP of JPG images.
- `ocr_pdf` adds searchable text to scans and accepts English, Russian, Spanish, Italian, Uzbek, or Kazakh language selection.
- `office_to_pdf` converts Word, PowerPoint, and Excel files to PDF.
- `image_to_pdf` places JPG, PNG, or WebP input into a PDF page.
- `webpage_to_pdf` renders a public URL with a headless browser.
- `protect_pdf` encrypts a PDF with AES-256, while `unlock_pdf` removes a password that is already known.

## Limitations and notes

OCR handles up to 15 pages per call. `webpage_to_pdf` requires an explicit `output_path` because it has no input filename from which to derive one. Guest access is rate-limited, so multi-step testing may exhaust the daily allowance. The service is intended for public webpage URLs, and all file-processing operations send the files to the Papyrio API.

_Full upstream README: https://allmcps.com/mcp/papyrio-pdf-tools/readme_