🔍 NewTel OCR Server

Intelligent Document Processing API

MCP Protocol v2024-11-05 Version 1.0.0

Overview

Welcome to NewTel OCR Server - a powerful, enterprise-grade document processing API built on the Model Context Protocol (MCP). Our service leverages advanced Mistral AI OCR technology to extract structured data from documents with unprecedented accuracy.

Designed for seamless integration, multi-tenant isolation, and scalable processing, this API is perfect for businesses that need reliable document analysis at scale.

Key Features

📄 Smart Document Processing

  • Automatic document type detection
  • Multi-language support
  • High-accuracy text extraction
  • Preserves document structure

📊 Advanced Data Extraction

  • Table recognition & parsing
  • List detection & structuring
  • Key-value pair extraction
  • Entity recognition (dates, amounts, emails)

🔒 Enterprise Security

  • Multi-tenant data isolation
  • API key authentication
  • Rate limiting & throttling
  • Secure file handling

⚡ Performance & Scalability

  • Fast processing times
  • Batch processing support
  • Result caching
  • Async job handling

🔄 Multiple Input Methods

  • Direct URL processing
  • Base64 encoded uploads
  • File upload via API
  • Local file path support

📈 Comprehensive Analytics

  • Processing statistics
  • Job history tracking
  • Full-text search
  • Export in multiple formats

Supported Document Types

PDF
Portable Documents
DOCX
Word Documents
XLSX
Spreadsheets
IMG
PNG, JPG, TIFF

API Endpoints

GET /ocr/mcp/health
Health check endpoint - verify server status and availability
POST /ocr/mcp
Main MCP endpoint - handles all document processing requests via JSON-RPC 2.0 protocol
POST /ocr/mcp/upload
File upload endpoint - upload documents for processing

Quick Start Example

Process a document using our API:

# Initialize MCP session curl -X POST https://bot.NewTel.eu/ocr/mcp \ -H "X-Api-Key: YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "jsonrpc": "2.0", "method": "tools/call", "params": { "name": "ocr_process", "arguments": { "source": { "type": "url", "data": "https://example.com/document.pdf" } } }, "id": 1 }'

Available Tools

ocr_process

Complete document processing with automatic data extraction including tables, lists, and entities.

ocr_extract_tables

Extract only tables from documents in structured JSON/CSV format.

ocr_analyze_document

Quick document analysis without full processing - detect type, language, and structure.

ocr_get_result

Retrieve processing results by job ID with flexible output formats.

ocr_search

Full-text search across all processed documents with advanced filtering.

ocr_list_jobs

List all OCR jobs with status filtering and pagination support.

Getting Started

To start using the NewTel OCR Server API:

  1. Contact your administrator to obtain an API key
  2. Review the MCP Protocol specification
  3. Test the health endpoint to verify connectivity
  4. Initialize an MCP session using the main endpoint
  5. Start processing documents with the available tools