EasyPDF

Official9 tools

by Lorenzino69

Edit PDF text in place, fonts and layout kept. Compress, merge, split, convert and translate PDFs.

MCP Server: io.github.Lorenzino69/easypdf

This MCP server exposes PDF operations for common document workflows. It can compress files, merge and split PDFs, rotate pages, extract pages, and number pages. It also supports securing documents by protecting PDFs or unlocking them as needed.

🛠️ Key Features

Compress PDFs
Merge PDFs
Split PDFs
Rotate pages
Extract pages
Number pages
Protect PDFs
Unlock PDFs

🚀 Use Cases

Reduce PDF size via compression
Combine multiple documents into one PDF
Divide a PDF into separate parts
Rotate pages to correct orientation
Extract specific pages for reuse
Add page numbering for readability
Secure documents or remove restrictions

⚡ Developer Benefits

Consolidates multiple PDF editing tasks (compress/merge/split/rotate/extract/number/protect/unlock) into one service.

Topics

chatgptclaudemcpmcp-appsmcp-servermodel-context-protocolpdfpdf-editor

Captured live from the server via tools/list.

edit_pdf_text

Change, correct or delete text INSIDE the original PDF, keeping its fonts, sizes, colors and layout. Use this whenever the user wants their existing PDF updated rather than a new document: fix a typo, change a date, amount, name or address, update a CV or a price list, translate a heading, remove a line. Read the PDF first, then pass one edit per change with `find` copied exactly as printed in the document. Dates and amounts written differently in the PDF (12/08/2026 vs August 12, 2026; 1240 vs 1,240.00) are matched and rewritten in the document's format.

Parameters2

file	object	required	The PDF to edit (shared in the conversation).
edits	array	required	Changes to write, applied in order. Each one replaces every occurrence of `find` (only on `page` when given).

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The PDF to edit (shared in the conversation).",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "edits": {
      "items": {
        "description": "One change to write into the PDF.",
        "properties": {
          "find": {
            "description": "Text to replace, copied exactly as it is printed in the PDF (same spelling, accents and punctuation). Keep it short: one line or a few words is most reliable.",
            "type": "string"
          },
          "replace": {
            "description": "New text written in place of `find`. An empty string deletes the text.",
            "type": "string"
          },
          "page": {
            "anyOf": [
              {
                "type": "integer"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "1-based page number to limit this change to. Omit to change every occurrence in the document."
          }
        },
        "required": [
          "find",
          "replace"
        ],
        "type": "object"
      },
      "type": "array",
      "description": "Changes to write, applied in order. Each one replaces every\noccurrence of `find` (only on `page` when given)."
    }
  },
  "required": [
    "file",
    "edits"
  ],
  "additionalProperties": false
}

compress_pdf

Compress a PDF to reduce its file size while keeping it readable.

Parameters2

file	object	required	The PDF to compress (shared in the conversation).
quality	string	optional	"low" (smallest file), "medium" (balanced, default) or "high" (best visual quality).

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The PDF to compress (shared in the conversation).",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "quality": {
      "default": "medium",
      "type": "string",
      "description": "\"low\" (smallest file), \"medium\" (balanced, default) or \"high\" (best visual quality)."
    }
  },
  "required": [
    "file"
  ],
  "additionalProperties": false
}

merge_pdfs

Merge several PDF files into a single document, in the given order.

Parameters1

files

array

required

Two or more PDFs shared in the conversation, in the order to merge them.

Raw schema

{
  "type": "object",
  "properties": {
    "files": {
      "items": {
        "description": "A file shared in the ChatGPT conversation (user upload).",
        "properties": {
          "download_url": {
            "description": "URL from which the file can be downloaded.",
            "type": "string"
          },
          "file_id": {
            "description": "ChatGPT file identifier.",
            "type": "string"
          },
          "mime_type": {
            "default": "",
            "description": "MIME type of the file, e.g. application/pdf.",
            "type": "string"
          },
          "file_name": {
            "default": "",
            "description": "Original file name, e.g. report.pdf.",
            "type": "string"
          }
        },
        "required": [
          "download_url",
          "file_id"
        ],
        "type": "object"
      },
      "type": "array",
      "description": "Two or more PDFs shared in the conversation, in the order to merge them."
    }
  },
  "required": [
    "files"
  ],
  "additionalProperties": false
}

split_pdf

Split a PDF into several smaller PDFs.

Parameters2

file	object	required	The PDF to split.
pages_per_file	integer	optional	Number of pages in each output file (default 1: one file per page).

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The PDF to split.",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "pages_per_file": {
      "default": 1,
      "type": "integer",
      "description": "Number of pages in each output file (default 1: one file per page)."
    }
  },
  "required": [
    "file"
  ],
  "additionalProperties": false
}

extract_pages

Extract specific pages from a PDF into a new PDF (pages are 1-indexed).

Parameters2

file	object	required	The source PDF.
page_numbers	array	required	Pages to keep, e.g. [1, 3, 5]. Order is preserved.

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The source PDF.",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "page_numbers": {
      "items": {
        "type": "integer"
      },
      "type": "array",
      "description": "Pages to keep, e.g. [1, 3, 5]. Order is preserved."
    }
  },
  "required": [
    "file",
    "page_numbers"
  ],
  "additionalProperties": false
}

rotate_pdf

Rotate all pages, or selected pages, of a PDF.

Parameters3

file	object	required	The PDF to rotate.
angle	integer	required	90, 180 or 270 degrees (clockwise).
pages	any	optional	Optional 1-indexed page numbers to rotate; omit to rotate every page.

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The PDF to rotate.",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "angle": {
      "type": "integer",
      "description": "90, 180 or 270 degrees (clockwise)."
    },
    "pages": {
      "anyOf": [
        {
          "items": {
            "type": "integer"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Optional 1-indexed page numbers to rotate; omit to rotate every page."
    }
  },
  "required": [
    "file",
    "angle"
  ],
  "additionalProperties": false
}

protect_pdf

Protect a PDF with a password (AES-256 encryption).

Parameters2

file	object	required	The PDF to protect.
password	string	required	Password that will be required to open the PDF.

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The PDF to protect.",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "password": {
      "type": "string",
      "description": "Password that will be required to open the PDF."
    }
  },
  "required": [
    "file",
    "password"
  ],
  "additionalProperties": false
}

unlock_pdf

Remove the password protection from a PDF you own.

Parameters2

file	object	required	The password-protected PDF.
password	string	required	Its current password.

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The password-protected PDF.",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "password": {
      "type": "string",
      "description": "Its current password."
    }
  },
  "required": [
    "file",
    "password"
  ],
  "additionalProperties": false
}

add_page_numbers

Add page numbers to a PDF.

Parameters3

file	object	required	The PDF to number.
position	string	optional	"bottom-center" (default), "bottom-left", "bottom-right", "top-center", "top-left" or "top-right".
start_number	integer	optional	Number printed on the first page. Default 1.

Raw schema

{
  "type": "object",
  "properties": {
    "file": {
      "description": "The PDF to number.",
      "properties": {
        "download_url": {
          "description": "URL from which the file can be downloaded.",
          "type": "string"
        },
        "file_id": {
          "description": "ChatGPT file identifier.",
          "type": "string"
        },
        "mime_type": {
          "default": "",
          "description": "MIME type of the file, e.g. application/pdf.",
          "type": "string"
        },
        "file_name": {
          "default": "",
          "description": "Original file name, e.g. report.pdf.",
          "type": "string"
        }
      },
      "required": [
        "download_url",
        "file_id"
      ],
      "type": "object"
    },
    "position": {
      "default": "bottom-center",
      "type": "string",
      "description": "\"bottom-center\" (default), \"bottom-left\", \"bottom-right\", \"top-center\", \"top-left\" or \"top-right\"."
    },
    "start_number": {
      "default": 1,
      "type": "integer",
      "description": "Number printed on the first page. Default 1."
    }
  },
  "required": [
    "file"
  ],
  "additionalProperties": false
}

README.md5 KBView on GitHub

EasyPDF MCP server

Your AI says what to change. EasyPDF writes it into the PDF.

Maintained by Lorenzino69, author of EasyPDF.

EasyPDF is a remote Model Context Protocol server that lets ChatGPT, Claude, Cursor, VS Code and any MCP client edit real PDF files. Ask your assistant to fix a date, an amount, a name or a typo: EasyPDF rewrites the text inside the original PDF, keeps the fonts and the layout, and shows a before/after preview in the chat. It also compresses, merges, splits, converts and translates PDFs.

https://github.com/user-attachments/assets/7eae0cd7-aea4-478e-b06c-91234c613fa4

21-second demo: one request, two edits written into the original invoice (fonts and layout kept), then the download. Also available as a GIF preview.

Website and setup guide: https://www.easypdf.fr/ai-assistants
Claude directory listing: https://claude.ai/directory/easypdf
No API key, nothing to install: it is a hosted server.

Endpoints

Client	URL	Authentication
Claude, Claude Code, Cursor, VS Code, other MCP clients	`https://www.easypdf.fr/mcp`	OAuth 2.1 (Google sign-in, dynamic client registration)
ChatGPT (apps / developer mode)	`https://www.easypdf.fr/chatgpt/mcp`	None

Transport: Streamable HTTP.

Install

Claude (claude.ai, desktop and mobile apps)

Open the listing at https://claude.ai/directory/easypdf, click Connect and confirm with your Google account.

Claude Code

The EasyPDF plugin adds the server plus skills that upload your local files and save the results next to them:

bash

claude plugin marketplace add Lorenzino69/easypdf-claude-plugin
claude plugin install easypdf@easypdf

Or the server alone. Keep the trailing slash: without it, Claude Code rejects the OAuth resource.

bash

claude mcp add --transport http easypdf https://www.easypdf.fr/mcp/

Cursor

~/.cursor/mcp.json

json

{
  "mcpServers": {
    "easypdf": { "url": "https://www.easypdf.fr/mcp" }
  }
}

VS Code

.vscode/mcp.json

json

{
  "servers": {
    "easypdf": { "type": "http", "url": "https://www.easypdf.fr/mcp" }
  }
}

ChatGPT

The EasyPDF app is under review for the ChatGPT app directory. Until then, add it in developer mode on chatgpt.com:

In Settings, open Apps, then Advanced settings, and turn on developer mode.
Create an app named EasyPDF with the URL https://www.easypdf.fr/chatgpt/mcp and no authentication.
In a chat, attach a PDF and pick EasyPDF from the + menu.

Tools

Tool	What it does
`edit_pdf_text`	Replaces text in the original PDF (dates, amounts, names, typos) with the same font, size and position. Returns a before/after preview widget (MCP Apps) and reports each replacement, with the closest match when a text is not found.
`get_upload_link`	Gives the user a one-hour upload link when the client cannot pass a file to the server.
`compress_pdf`	Reduces the file size.
`merge_pdfs`	Merges several PDFs in order.
`split_pdf`	Splits a PDF into files of N pages.
`extract_pages`	Keeps only the listed pages.
`rotate_pdf`	Rotates all or some pages by 90, 180 or 270 degrees.
`add_page_numbers`	Adds page numbers.
`watermark_pdf`	Adds a text watermark.
`protect_pdf`	Adds a password.
`unlock_pdf`	Removes a known password.
`convert_pdf_to_word`	Converts to .docx.
`convert_pdf_to_excel`	Converts to .xlsx.
`convert_pdf_to_powerpoint`	Converts to .pptx.
`translate_pdf`	Translates the document and keeps the layout.
`generate_pdf`	Creates a PDF from a text prompt.
`chat_with_pdf`	Answers a question about a PDF, with source references.

The ChatGPT endpoint exposes the tools that work on ChatGPT attachments: edit_pdf_text, compress_pdf, merge_pdfs, split_pdf, extract_pages, rotate_pdf, add_page_numbers, protect_pdf and unlock_pdf.

Every tool returns a download link valid for one hour.

Example prompts

"In this invoice, change the due date 17/10/2026 to 31/10/2026 and keep the layout."
"Fix the spelling mistakes in my resume directly in the PDF."
"Replace "Example Ltd" with "Example Consulting Inc." everywhere in this contract."
"Compress this PDF under 2 MB so I can upload it to an online form."

Privacy

Uploaded PDFs are used during processing and deleted within one hour.
Output files are available through a download link for one hour, then deleted.
Text edits run without any AI model on the EasyPDF side: your assistant decides what to change, EasyPDF applies the replacements exactly.
EasyPDF receives the file and the tool arguments, never the rest of your conversation.

Support

Questions or bugs: open an issue in this repository or write to easypdfbug@gmail.com.

This repository documents the hosted EasyPDF MCP server (the server itself is closed source). Documentation is released under the MIT license.

Install

Tools (9)

edit_pdf_textcompress_pdfmerge_pdfssplit_pdfextract_pagesrotate_pdfprotect_pdfunlock_pdf

Remote endpoint

Streamable HTTP

Hosted server - connect over the network, no local install.

url

https://www.easypdf.fr/mcp

Add to VS Code Add to Cursor

claude_desktop_config.json

json

{
  "mcpServers": {
    "easypdf": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://www.easypdf.fr/mcp"
      ]
    }
  }
}

Desktop config is stdio-only; this bridges via mcp-remote. Native remote: Settings > Connectors.

MCP Server: io.github.Lorenzino69/easypdf

🛠️ Key Features

🚀 Use Cases

⚡ Developer Benefits

Topics

edit_pdf_text

compress_pdf

merge_pdfs

split_pdf

extract_pages

rotate_pdf

protect_pdf

unlock_pdf

add_page_numbers

EasyPDF MCP server

Endpoints

Install

Claude (claude.ai, desktop and mobile apps)

Claude Code

Cursor

VS Code

ChatGPT

Tools

Example prompts

Privacy

Support

Install

Remote endpoint

Related servers