A reading library of its own: 22,081 books, transcribed and translated, open to anyone.
Ancient, hermetic and early modern texts drawn from open archives worldwide, in Latin, Greek, Tibetan, Ge’ez, Chinese and Yucatec Maya, published as a single catalogue.
It shows the process working on difficult material at volume: early type with long s and ligatures, scripts that run right to left, and manuscripts transcribed for the first time.
Search a phrase in English and arrive at the line in the original. Read the translation beside the scan, with the transcription between them, line for line.
Follow an illustration out of the book into a gallery of related images. Every scanned page is read to find, extract, tag and caption the illustrations, figures and diagrams buried in the texts, so an image you could never have searched for becomes findable in seconds.
Ask a question in plain language and the librarian finds, summarises and cross-references across thousands of translated sources. It is open to machines too: an MCP server, a CLI and an API let other AI tools read the library directly.
“Thousands of these works are available in a modern reading language for the first time, and a student anywhere can open them tonight.”sourcelibrary.org
Send a sample and whatever catalogue export you have. We publish it properly and show you the result on real pages.
