Receipt Scanner

An Apify actor for automatically extracting structured data from receipt and invoice images using AI. It processes image URLs through OpenAI or OpenRouter models and returns parsed data in JSON.

Note: It is designed exclusively for receipts and invoices. Processing other document types is not guaranteed. Input must be an image URL; direct file upload is not supported.

Key Features

• AI-Powered Extraction: Uses state-of-the-art models from OpenAI or OpenRouter for high-accuracy text and structure recognition.
• Structured Data: Extracts key fields like store, date, line items, totals, taxes, and currency.
• Enhanced Receipt Data: Also captures specialized fields such as:
  • Loyalty program information and points earned.
  • Return policy details.
  • Warranty information.
  • Gift receipt indicators.
  • Promotional codes.
• Flexible & Scalable: Supports both single receipt and batch processing with configurable concurrency.
• Caching: Implements result caching to speed up repeated processing of the same image.
• Easy Integration: Offers a simple Python interface and an API for quick embedding into business processes.

How to Use

Setup

1. Install the required Python package:
   bash pip install -r requirements.txt
2. Configure environment variables for your chosen AI provider and optional caching:
   bash export OPENAI_API_KEY=your_key export OPENROUTER_API_KEY=your_alternative_key # Optional caching export REDIS_URL=redis://localhost:6379/0 export CACHE_TTL_SECONDS=86400

Python Interface

Process a Single Receipt URL

from src.utils import process_receipt_url
result = process_receipt_url("https://example.com/receipt.jpg")

Batch Process Multiple URLs

from src.utils import process_receipt_urls

urls = ["https://example.com/receipt1.jpg", "https://example.com/receipt2.jpg"]
results = process_receipt_urls(urls, max_concurrent=5)

for result in results:
    if 'data' in result:
        data = result['data']
        print(f"Store: {data.get('store', {}).get('name')}")
        print(f"Total: {data.get('totals', {}).get('total')}")
        # Access enhanced data
        if 'loyalty' in data:
            print(f"Points Earned: {data['loyalty'].get('points_earned')}")
    elif 'error' in result:
        print(f"Error: {result['error']}")

API Usage

The actor provides a web API endpoint. After deployment on the Apify platform, you can send POST requests with a JSON payload containing image URLs.

Input & Output

• Input: One or more publicly accessible URLs pointing to receipt/invoice images (JPG, PNG, etc.).
• Output: A standardized JSON structure containing the extracted data, the source URL, and a cache status flag. Errors for failed processing are included in the response.

Costs

The service is free during testing. Future costs will follow Apify's pricing. The final cost depends on image volume/complexity and the chosen AI provider/model.

• Apify Platform: Costs are per compute unit. See Apify pricing page.
• AI Provider Fees: You incur separate costs from your chosen AI provider.
  • OpenAI: Billed per token according to their official pricing.
  • OpenRouter: An alternative provider; see openrouter.ai/pricing.