How it works
AI trained on five centuries of Spanish handwriting
Transkribus uses deep learning trained on millions of handwritten pages, from the 1400s to the present. The AI has learned the distinctive forms of old Spanish scripts – the tangled ligatures of Procesal, the angular strokes of Cortesana, the flowing lines of colonial-era letters – and turns them into clean, searchable text you can read and export.
Reads Procesal, Cortesana, Encadenada, Humanística, and Itálica scripts
300+ public models covering different handwriting styles and centuries
Works with phone photos, archival microfilm scans, or high-resolution digitisations
What you can transcribe
Documents from Spain, the Americas, and the Philippines
Your document might be a parish register from Mexico, a notarial protocol from Seville, or a land grant from colonial Peru – Spanish-language records are scattered across dozens of countries, and almost everything before 1900 is handwritten. Transkribus handles the handwriting so you can focus on the content.
Parish registers (registros parroquiales) – baptisms, marriages, and burials across Spain and Latin America
Notarial records (protocolos notariales) – contracts, wills, dowries, and property transfers
Inquisition proceedings and trial records
Ship manifests and passenger lists from the Carrera de Indias
Census records (padrones) and tax rolls
Viceregal correspondence, royal decrees, and land grants (mercedes de tierra)
Beyond the demo
The full platform for large-scale projects
The demo above handles one page at a time. The full Transkribus platform lets you upload entire archive boxes, process thousands of pages in a single batch, train a custom AI model tuned to your specific collection's handwriting, and run full-text search across everything you have transcribed.
Batch-process hundreds or thousands of pages at once
Train custom models on your archive's particular scribal hands
Full-text search across all transcribed documents in your workspace
Export as TXT, DOCX, PDF, TEI-XML, or PAGE XML
Genealogy and research
Common documents in old Spanish handwriting
Whatever record you're chasing, it was probably written in a script that's been out of everyday use for centuries. Colonial-era records are especially tough: many are in Procesal or similar notarial hands that even experienced researchers struggle to read.
Parish vital records from Mexico, Peru, Colombia, the Philippines, and across Latin America
Protocolos notariales – the most common source for genealogy before civil registration
Military service records, enlistment rolls, and campaign correspondence
Diligencias matrimoniales – marriage investigation files with witness testimony
Catastro de Ensenada and other property and census surveys
Personal letters, diaries, and household account books
Which script is it?
Found a Spanish document you can't read? Here's what you're probably looking at.
The script depends on when and where your document was written – a notarial act, a parish entry, and a personal letter from the same decade can look completely different. These are the most common old Spanish scripts:
Procesal – the notarial hand of the 1500s and 1600s, in Spain and across Latin America. Dense abbreviations and tangled ligatures make it notoriously hard to read. If your document looks like an unbroken scrawl, it's probably this.
Cortesana – a rounder hand in royal and administrative documents from the 1400s and 1500s. Easier to read than Procesal.
Procesal encadenada – an extreme form of Procesal where whole words are chained together without lifting the pen. The hardest of all.
Humanística – round, clear letters in books and formal texts. If you can make out most words, it's probably this.
Itálica – the sloped cursive that took over from the 1700s onward. Documents from then on look much closer to modern handwriting.
The technology
How does Spanish handwriting recognition work?
Handwritten Text Recognition (HTR) uses deep learning to convert images of handwritten text into machine-readable characters. This is much harder than printed-text OCR because every writer forms letters differently, abbreviations vary by region and century, and scripts like Procesal compress multiple characters into single pen strokes that a human eye must learn to unravel.
Neural networks trained on labelled samples of historical Spanish handwriting
Automatic layout analysis detects text lines, columns, and marginal notes
Character-level recognition handles the heavy ligatures and abbreviations typical of Procesal
Language models use context to disambiguate where the handwriting alone is unclear
AI Models for Spanish Handwriting
Browse public models trained on historical Spanish documents – Procesal, Cortesana, colonial-era hands, and more.
Ready to read old Spanish handwriting?
Create a free account to process unlimited documents, train custom models, and unlock the full platform. Want a quick test? Try the free handwriting translator – no signup needed.
50 free credits every month – No credit card required
200M+Pages processed
500K+Users worldwide
300+Public AI models