Amazon Textract

API

AWS API that reads text, forms, tables and invoices from scanned documents

Price: 3-month free trial, then pay per use
Access: AWS credentials

About

AWS service that detects printed and handwritten text and analyzes forms, tables, invoices, receipts, US IDs and mortgage packages. Synchronous calls take one page; multipage PDFs go through S3 with async jobs. Text detection covers English and five other European languages.

What you can do with it

Detect printed and handwritten text in scanned PDFs and images
Extract forms, tables and answers to natural-language queries from documents
Read vendor, totals and line items from invoices and receipts

Get started

Create an IAM user with Textract access and an access key
Configure the AWS CLI with that profile
Upload the document to an S3 bucket in the same region

Example

aws textract detect-document-text \
  --document '{"S3Object":{"Bucket":"my-bucket","Name":"scan.png"}}' \
  --region us-east-1

# Invoices and receipts
aws textract analyze-expense \
  --document '{"S3Object":{"Bucket":"my-bucket","Name":"invoice.pdf"}}' \
  --region us-east-1

Details

Hosting: Hosted service
Available in: Worldwide
Official SDKs: Python, JavaScript/TypeScript, Java, Go, Ruby, PHP, C#, Kotlin, Rust, Swift
MCP server: None
Works with: AWS, S3