Global Spirits Inventory

Country is the top axis, spirit type the next. Each row records the distillery that made the liquid rather than the brand on the label, where it was distilled, the strength, the recipe, the cask, and any age statement or named release, across twenty-one categories from bourbon and Scotch to mezcal, armagnac and vermouth.

The part that took the work is not the size. Every row carries the URLs that substantiate it and the date those sources were last checked, so any entry can be traced back and re-verified rather than taken on trust. Plenty of large lists exist. Very few can tell you where a number came from.

Get the data

The repository holds the whole set as one CSV, and the same rows split by spirit type and region for anyone who wants a single category. A build manifest records the row counts and the SHA-256 of the consolidated file, so a copy can be checked against the build it came from.

Published under CC BY 4.0. Free to use, adapt and build on commercially, with attribution the only condition. Corrections by pull request or issue, and they are welcome: the people who know a given category best are rarely the people who compiled it.

What it does not do yet

The repository ships a KNOWN_GAPS file rather than a claim of completeness. 12,889 rows compiled from the public web contain mistakes, and the ones already found are named there: rows citing a producer homepage instead of the page for that bottling, a column whose meaning is inconsistent between categories, and a handful of entries whose age statement does not square with when their distillery opened. Tasting notes are withheld, because the text is producer copy and this licence cannot grant rights to words the project does not hold.

A quarterly pass adds new releases and works down that list. Each pass is tagged, so a citation points at data that does not move.