Extracting data from a PDF is now a single API call
DocuTray is the document processing API for developers — trained for LATAM documents.
# paso 1 — envía el documento a /api/convert curl -X POST https://app.docutray.com/api/convert \ -H "Authorization: Bearer $DOCUTRAY_API_KEY" \ -F "image=@cartola-tc-marzo.pdf" \ -F "document_type_code=cartola_tarjeta_credito" # también acepta JSON con base64 o URL. # identificar → convertir → validar → exportar.
// beyond OCR: the full flow in one call
Clientes que confían en DocuTray
A developer experience as direct as an API call
Any document, any format, any layout. In seconds.
Multiple formats
Extract data from PDFs, images, photos and scanned documents. Any format, any layout.
AI accuracy
OCR + AI specialized and tuned for Spanish documents: high accuracy even on tough scans.
Blazing fast
Process hundreds of documents in minutes with our high-performance extraction engine.
Security & privacy
High security standards and configurable data retention for your confidential information.
API integration
REST + webhooks. Upload via base64, form-data or URL and integrate into your systems in minutes.
Custom documents
Build extractors for your organization’s internal documents with the help of our assistant.
Document processing is a mess. Even with AI.
Your own OCR pipeline
- ✕Choosing and orchestrating OCR, LLMs and RAG
- ✕Cloud infrastructure and scaling
- ✕Error handling and retries
- ✕Monitoring, logging and updates
- ✕Custom validations per document type
- ✕Security and compliance
…instead of focusing on your product.
One call. Clean JSON.
- ✓A single API call
- ✓No infrastructure to maintain
- ✓Integrate in minutes, not weeks
How DocuTray
Six stages of intelligent document processing. One single call — scroll to walk through them.
API Call
You send the document —PDF, image or scan— in a single request. Pick whatever method suits you: file (form-data), base64 or a URL — they all hit the same endpoint.
# multipart/form-data — subida directa del archivo curl -X POST https://app.docutray.com/api/convert \ -H "Authorization: Bearer •••" \ -F "image=@cartola-tc-marzo.pdf" \ -F "document_type_code=cartola_tarjeta_credito"
How DocuTray
Six stages of intelligent document processing. One single call — scroll to walk through them.
API Call
You send the document —PDF, image or scan— in a single request. Pick whatever method suits you: file (form-data), base64 or a URL — they all hit the same endpoint.
# multipart/form-data — subida directa del archivo curl -X POST https://app.docutray.com/api/convert \ -H "Authorization: Bearer •••" \ -F "image=@cartola-tc-marzo.pdf" \ -F "document_type_code=cartola_tarjeta_credito"
Identify
DocuTray automatically recognizes which document it is. You don’t have to tell it or maintain rules per format.
// webhook IDENTIFICATION_COMPLETED { "status": "SUCCESS", "document_type": { "code": "cartola_tarjeta_credito", "name": "Cartola tarjeta de crédito", "confidence": 0.98 }, "alternatives": [ { "code": "cartola_cuenta_corriente", "confidence": 0.12 } ] }
Convert
OCR + AI read the content and structure it into the JSON schema you defined, even with tough scans or photos.
"transacciones": [ { "fecha": "2026-03-04", "comercio": "LIDER", "monto": 45990 }, { … } // +84 movimientos ]
Validate
Business rules catch inconsistencies before they reach your system: balance checks, totals and formats.
// webhook STEP_COMPLETED → validation "validation": { "errors": { "count": 0, "messages": [] }, "warnings": { "count": 1, "messages": ["Monto supera el promedio histórico"] } }
Enrich
We cross the data with your internal databases or external sources to complete what the document doesn’t carry.
{
"comercio": "LIDER",
"monto": 45990,
"categoria": "supermercado", // ← enriquecido
"rubro_sii": "471000" // ← enriquecido
}Export
We deliver the result to your systems via webhook in real time, or you download it as JSON or Excel.
→ POST https://tu-app.com/webhooks/docutray { "status": "completed", "export": ["json", "xlsx"] }
Trained on the documents used across LATAM
OCR and data extraction in Spanish: from bank statements to SII and SAT invoices, including foreign-trade documents.
Checking account statement
Transactions, balances and holders from any bank statement.
Credit card statement
Domestic and international transactions, limits and payments.
Investment statement
Positions, movements and returns from brokers and funds.
SII invoices
Chilean invoices, credit notes and fee receipts.
SAT invoices
Mexican CFDI and tax receipts.
International invoicing
Commercial invoices from foreign suppliers, in any currency.
Packing list
Packages, weights and contents, ready for your system.
Bill of Lading / Waybill
Bills of lading and transport documentation.
Medical prescription
Reads prescriptions, even with doctors’ handwriting.
Your document isn’t on the list? Build a custom extractor with our AI →
What our customers say
“We were looking for an OCR provider, and DocuTray ended up being much more than that. With a single API call, we can identify the document, extract the relevant data and validate its consistency, achieving results far superior to other solutions we tried.”
“DocuTray’s API gave us a superpower: we can automatically read medical prescriptions, even with doctors’ handwriting. Now we can build experiences for our customers that we couldn’t before.”
Compare DocuTray with other solutions
We replace the entire document processing flow — not just the OCR.
Ready to extract data from your documents?
Start processing documents in minutes, not weeks. Create your free account or write to us: we reply the same day.