Glintstone

Five thousand years of writing.
Less than 2% translated.

The unified cuneiform research infrastructure. Every artifact. Every annotation. Every scholar. Fully attributed.

Scale. Provenance. Open.

Aggregated open scholarship from the world's leading cuneiform institutions. Every record carries its source, license, and provenance.

0
Artifacts
0
Translations
0
Lemmatizations
0
Scholars
CDLI ORACC eBL ePSD2 OGSL Pleiades

Open data from the world's leading cuneiform research institutions. Every record carries its source, license, and provenance.

Built for scholars who care about attribution

Every annotation in Glintstone — transliteration, lemmatization, translation, AI suggestion — carries who made it, when, and how. Competing interpretations are both stored. No reading overwrites another.

When two scholars disagree on a word, Glintstone keeps both readings, both attributed. When a machine learning model proposes a translation, it is labeled as such, with its confidence score and the evidence it was given.

This is infrastructure for the scholarly record, not a replacement for it.

Research infrastructure, not a search box

Explore the unified archive

Search 353k cuneiform artifacts across CDLI, ORACC, and eBL. Filter by period, genre, language, provenience. See exactly what is known — and what remains unknown.

Track pipeline completeness

Every artifact moves through five stages — captured, recognized, transcribed, lemmatized, translated. See at a glance where scholarly effort has the highest leverage.

AI translations with grounded citations

AI-assisted translations with a citation for every claim. See the full token chain. Know when scholars disagree. Know when the model is estimating.

Early access researchers only

Request early access

Glintstone is in early access for academic researchers. We are looking for Assyriologists, philologists, and computational linguistics researchers to help shape what this becomes.

We'll respond within 2 weeks. No spam.