← Back to Blog

Business Card Data Extraction API Guide

tutorialaiapibusinesscarddata

Business Card Data Extraction API: A Complete Developer Guide

Manually typing contact details from business cards is one of those small tasks that quietly eats hours every week. Whether you’re processing leads after a trade show or building a CRM integration, a business card data extraction API can turn a stack of scanned cards into structured, usable data in seconds.

This guide walks you through how the Today’s World AI API handles business card OCR, what the request and response structure looks like, and how to integrate it into a real workflow.

What Is a Business Card Data Extraction API?

A business card data extraction API accepts an image or base64-encoded scan of a business card and returns structured contact data — names, job titles, phone numbers, email addresses, company names, and more. Under the hood, it combines optical character recognition (OCR) with natural language processing to intelligently parse unstructured card layouts.

This is different from a generic OCR tool. A business card OCR API understands context. It knows that a string like j.smith@acmecorp.com is an email, not a job title — even when the card layout is unconventional or multilingual.

Why Automate Business Card Digitization?

If you’re building or scaling any of the following, vCard automation and contact data extraction become essential:

  • CRM pipelines — Auto-populate Salesforce, HubSpot, or Zoho from scanned cards
  • Event management apps — Let attendees scan and exchange cards digitally
  • Lead capture tools — Convert physical cards to digital contacts in the field
  • Document processing platforms — Bundle card scanning alongside invoice or receipt parsing

If you’re also handling other document types, you might find our related posts useful: How to Extract Invoice Data from PDFs Using AI and Parse Receipts Automatically with AI API cover similar extraction patterns for financial documents.

Getting Started with the API

Before making your first call, you’ll need an API key. You can get started immediately with a free tier, or browse the full reference in the API documentation.

Prerequisites

  • An active Today’s World AI API key
  • An image of a business card (JPEG, PNG, or PDF)
  • Basic familiarity with HTTP requests or a REST client like Postman

Making Your First Request

The digitize business cards API endpoint accepts a POST request with either a publicly accessible image URL or a base64-encoded image string.

Request Format

Here’s a working example using a URL-based image input:

curl -X POST https://api.todaysworld.com/v1/business-card/extract \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "image_url": "https://example.com/cards/john_doe_card.jpg",
    "output_format": "structured",
    "include_vcard": true,
    "language": "en"
  }'

Request Parameters

ParameterTypeRequiredDescription
image_urlstringYes*Publicly accessible URL of the card image
image_base64stringYes*Base64-encoded image (alternative to image_url)
output_formatstringNostructured (default) or raw
include_vcardbooleanNoReturns a vCard 3.0 string if true
languagestringNoISO 639-1 language code (default: en)

*Provide either image_url or image_base64, not both.

Understanding the Response

A successful request returns a JSON object with all extracted fields clearly labeled. Here’s an example response:

{
  "status": "success",
  "confidence": 0.97,
  "data": {
    "full_name": "Jane Merritt",
    "first_name": "Jane",
    "last_name": "Merritt",
    "job_title": "Senior Product Manager",
    "company": "Acme Corporation",
    "email": "jane.merritt@acmecorp.com",
    "phone": [
      { "type": "work", "number": "+1 (415) 555-0192" },
      { "type": "mobile", "number": "+1 (415) 555-0847" }
    ],
    "website": "www.acmecorp.com",
    "address": {
      "street": "350 Market Street",
      "city": "San Francisco",
      "state": "CA",
      "zip": "94105",
      "country": "US"
    },
    "linkedin": "linkedin.com/in/janemerritt"
  },
  "vcard": "BEGIN:VCARD\nVERSION:3.0\nFN:Jane Merritt\nORG:Acme Corporation\n..."
}

Key Response Fields

  • confidence — A score between 0 and 1 indicating extraction reliability. Scores above 0.85 are generally safe for automated processing.
  • data — The fully structured contact object.
  • vcard — Only present when include_vcard: true. This string can be imported directly into most contact applications and CRMs.

Handling Edge Cases

Real-world business cards are messy. Here are a few things to keep in mind:

Low-Quality Images

The API performs best with images that are at least 300 DPI and well-lit. If confidence scores are consistently below 0.75, consider adding an image pre-processing step (contrast enhancement, deskewing) before sending the request.

Multiple Languages

Set the language parameter to match the card’s primary language. Cards with mixed-language content (common in international business settings) will still extract correctly in most cases, but explicit language hints improve accuracy.

Missing Fields

Not all cards include every field. Missing values are returned as null rather than empty strings, making it straightforward to handle them in your application logic.

Integrating into a CRM Workflow

A common use case is combining this API with a webhook or automation tool like Zapier or Make. Here’s a simplified flow:

  1. User uploads a card photo through your mobile app
  2. Your backend sends the image to the business card data extraction API
  3. The structured response is mapped to CRM fields
  4. A new contact record is created automatically
  5. Optional: the vCard string is also emailed to the user as a backup

For teams processing high card volumes, batch endpoints and asynchronous processing options are documented in the full API docs.

Pricing and Rate Limits

The free tier includes up to 100 extractions per month — enough to prototype and test your integration thoroughly. Paid plans scale based on monthly volume, with no per-field charges. All tiers include access to the vCard automation output and multilingual support.

Summary

A business card OCR API removes a surprisingly stubborn bottleneck from contact management workflows. With structured JSON output, built-in vCard support, and high confidence scoring, the Today’s World AI business card data extraction API is straightforward to integrate whether you’re building a standalone scanning app or adding a feature to an existing platform.

Check the API documentation for advanced options including webhook callbacks, batch processing, and SDK examples in Python, Node.js, and PHP.

Ready to automate your workflow? Try it free at todaysworld.com/try or get API access on RapidAPI.

Try Free → API Docs →