mcpdirectory.dev
9,129 servers · updated 23 Sept

Ocr MCP server

community Active last commit 53 d
{
  "mcpServers": {
    "auto-reader-ocr": {
      "type": "http",
      "url": "https://api.auto-reader.com/mcp"
    }
  }
}
file .mcp.json (in your project root)version 1.0.3

What it does

Extracts text from images and PDFs with support for Arabic, Japanese, and 13+ languages, translates content across 100+ languages, and pulls structured data from documents like invoices and IDs. Designed for agents and AI applications that need OCR, translation, and document parsing in a single API.

  • Extract Arabic text from scanned invoices and validate Saudi ZATCA e-invoice QRs for compliance.
  • Convert document images into RAG-ready markdown or semantic chunks for retrieval-augmented generation systems.
  • Extract structured fields like names, amounts, and dates from receipts, invoices, and ID documents with confidence scores.

Tools this server exposes

ToolDoes
ocr_imageExtract text from an image with auto-detected language support for Arabic, Japanese, and 13+ languages.
extract_documentPull structured fields from invoices, receipts, and ID documents with confidence scores and bounding box information.
translate_textTranslate text to or from 100+ languages.
ocr_and_translatePerform OCR on an image and translate the extracted text in a single call.
get_usageCheck current usage statistics and remaining quota.
create_api_keyGenerate a new API key for authentication.

Before you enable it

An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.