Skip to main content
Identification happens in two steps. The API gets you to the album right away. Getting to the exact pressing is optional, and your app decides how far to go.

1. Get the album

Barcode (best)

Scan the barcode on the back. It’s the most reliable input: 99.2% of Discogs vinyl barcodes belong to exactly one album, and 83.4% to exactly one pressing.
Pass the barcode as printed. Spaces, dashes and leading zeros don’t matter, and UPC-12 and EAN-13 forms of the same code match each other. The response’s match field tells you what to do next:
Decode the barcode on the device (for example with the browser’s BarcodeDetector API) and send only the digits. It’s faster and no image leaves the phone.

Artist and title

Word order doesn’t matter and “the” is ignored. Albums with many vinyl versions rank above one-off releases with similar names.

2. Narrow to the pressing

With filters and facets

GET /v1/albums/{id}/pressings filters the album’s pressings and returns facets: the count of each remaining value. Ask the user about the facet that splits the list best.
The Dark Side of the Moon goes from 674 pressings to 111 with country=US, then to 29 with label=Harvest&year=1973.

With codes from the record

The runout etching (the text scratched into the vinyl near the label) is the strongest pressing identifier. Any fragment works:
/v1/search also takes catno, barcode, label, country, artist, title and year, and scores each candidate by which fields matched. When pressings tie, note says what would separate them.

What separates tied pressings

In testing, pressings that tie on every code usually differ only in things no code captures: vinyl color, a limited edition, an insert, or the “Made in…” text on the back cover. Ask the user, or read it from a photo of the label.

Example: four real records