Discogs Scraper: releases, barcodes and collector stats

Discogs Scraper searches releases, masters, artists and labels and returns catalog rows with barcodes, formats and collector statistics. Search without a Discogs account or add your personal token. It works like a no-code Discogs API: fill in a search, get a dataset you can export as JSON, CSV or Excel.

Use it to match a barcode from your inventory to a specific pressing, look up releases by artist or label, or collect catalog records for further analysis. Every result includes the original Discogs identifier and a link to its page. The actor searches the Discogs database; it does not return marketplace listings, asking prices, transaction history or estimates of sales.

Discogs Scraper input

Provide at least one search field. All fields are optional individually, so you can search by barcode alone, by artist alone, or combine several filters. A run without any search field returns no rows and an input explanation.

Field	Default	Purpose
q	No default; the example is nirvana	Free-text catalog search
type	release	release, master, artist or label
artist	Empty	Artist-name filter
label	Empty	Label-name filter
barcode	Empty	EAN or UPC lookup
catno	Empty	Catalog-number filter
genre, style	Empty	Discogs genre and style filters
country, year, format	Empty	Pressing country, year and format filters
track	Empty	Track-title filter
token	Empty	Optional personal Discogs access token
limit	100	Maximum results, from 1 to 2000
retries	3	Attempts per page, from 1 to 6

The original additional API fields title, release_title, credit, anv, submitter and contributor remain supported. year accepts a string or an integer. Input values are trimmed before they are sent to Discogs. country filters the release country recorded in the catalog, rather than the location of the buyer.

Here is the actual input used for the barcode example below:

{"barcode": "602547378781", "limit": 100}

This barcode returned four catalog entries in the recorded search. Barcode matches can include more than one pressing, so inspect the identifier, catalog number and other metadata before joining a result to your stock. A shared barcode alone does not establish that two physical copies are the same edition.

Output

Results appear in the default dataset, one row per unique Discogs id. position is one-based across the entire run. The following complete row was returned by a real anonymous search on 3 October 2026:

{
  "id": 29106535,
  "type": "release",
  "title": "Nirvana - Nirvana",
  "year": "2015",
  "country": "Europe",
  "url": "https://www.discogs.com/release/29106535-Nirvana-Nirvana",
  "labels": [
    "DGC",
    "Sub Pop",
    "Geffen Records",
    "Geffen Records"
  ],
  "catalog_number": "0602547378781",
  "barcode": "602547378781",
  "barcodes": [
    "602547378781"
  ],
  "formats": [
    "Vinyl",
    "LP",
    "Compilation",
    "Reissue"
  ],
  "genres": [
    "Rock"
  ],
  "styles": [
    "Grunge",
    "Alternative Rock"
  ],
  "master_id": 42476,
  "want": 131,
  "have": 4587,
  "want_have_ratio": 0.029,
  "thumbnail": null,
  "cover_image": null,
  "position": 1,
  "source": "discogs-scraper",
  "platform": "discogs",
  "entry_id": 29106535,
  "entry_type": "release",
  "gtin": "602547378781",
  "image_url": null,
  "fetched_at": "2026-10-03T14:45:22.340Z"
}

All original result fields remain available. year stays a string when known. labels retains repeated label credits, and formats, genres and styles are lists. Missing values are null, rather than guessed. New fields provide a shared catalog vocabulary: source, platform, entry_id, entry_type, gtin, image_url and fetched_at. The original id, type, barcode and cover_image remain alongside those fields.

The barcode field is the first cleaned barcode, while barcodes contains all distinct cleaned values. Discogs can mix matrix strings and rights-society labels into its barcode array. The actor removes non-digit characters and keeps only entries containing 8 to 14 digits. This preserves usable catalog join keys without treating every identifier as a retail barcode.

want counts Discogs users who have added a pressing to their wantlist; have counts the collections recorded by that community. want_have_ratio divides want by have and rounds to three decimal places. It is null if the necessary counts are unavailable or have is zero. These fields describe Discogs users, not worldwide demand, actual transactions, revenue or market share.

Run summary and errors

The default key-value store contains summary, with the delivered item count, total matches, pages fetched, barcode coverage, community-stat coverage and elapsed fetch time. A samplingNote explains when the search returned only part of the matching catalog. Compare resultsReturned with totalMatches before treating a search as exhaustive.

The errors record is written when a problem occurs. A search with no matches finishes successfully with an empty dataset and a no_results explanation. Invalid search input, a rejected personal token and a temporary source failure also produce readable explanations. If a later page fails, rows already collected remain in the dataset and the summary identifies the partial result. Internal processing of larger searches is automatic in the actor.

Pricing

There are two existing events: run_start costs $0.001 once per run, including runs that return nothing; record costs $0.001 for each result written to the dataset. The record rate is $1 per 1,000 delivered rows. Empty searches incur the start fee and no record charges. Duplicate ids do not create extra dataset rows or extra record charges.

For users without a paid Apify plan, each run delivers at most ten rows. Paid users retain their requested limit, up to 2,000. An interrupted or unavailable page does not generate result charges for rows that were never delivered. Export the dataset in the format you need through Apify's normal dataset tools.

Example: a paid user who runs a barcode search that returns 4 rows pays $0.001 for the start and 4 × $0.001 for the rows, $0.005 in total. A label search with a limit of 1,000 that delivers 1,000 rows costs $1.001.

Use cases

• Record shops and resellers: match the barcodes in your stock list to Discogs releases and pull in the catalog number, label, country and format for your product pages.
• Catalog enrichment: add Discogs ids, master ids, genres and styles to an existing music catalog so related pressings can be grouped.
• Collection and demand research: compare want and have across the pressings of an album or a label to see which editions collectors are chasing.
• Label and artist discographies: export every release, master or label entry for an artist or label filter in one run instead of paging through the website.
• Data pipelines: schedule the actor on Apify and send the dataset to a spreadsheet, a database or your own code through the Apify API.

Discogs Scraper compared with the alternatives

Option	What you get	What it takes
Discogs website search	The same catalog, one page at a time	Manual copying, no export
Official Discogs API	Full access for developers	Your own code, pagination, rate-limit handling and retries
Discogs monthly data dumps	The whole database as large XML files	Downloading and parsing many gigabytes; no collector counts
Other Store actors	Varies; several also scrape marketplace prices	Check whether their rows include barcodes and collector stats
This Discogs Scraper	Clean rows with barcodes, formats, master id and want/have counts	A search form or a JSON input, nothing to host

This actor is a search tool for the catalog. If you need marketplace asking prices, a marketplace-focused actor is the better fit; if you need barcodes, pressing details and collector demand per release, this one returns them in one row.

Uses and limits

Use type: release when you need individual pressings with country, format and catalog-number metadata. Use type: master when you need the grouping record across different pressings. Artists and labels use the same row structure, but release-specific fields can be null. A master id on a release helps you group related pressings for later comparison.

For a broad label or genre search, set an explicit limit and inspect the match count. The actor reads at most twenty pages of one hundred source results. The result list is deduplicated by id, so changes or overlap between source pages can reduce the delivered count below the requested limit. Narrowing the search is useful when the reported match count is much larger than the rows you need.

Image fields may be null in anonymous responses. A personal Discogs token can make authenticated metadata available, but the actor does not invent missing images or promise that every result includes one. Tokens are used for Discogs authentication and are excluded from result rows, error messages and summaries.

FAQ

Do I need a Discogs account?

No. Anonymous catalog searches are supported. You can optionally provide a personal Discogs token. If it is rejected, remove it or replace it with a valid personal token and run the search again.

Can I identify a pressing by barcode?

Yes. Use barcode to search by EAN or UPC, then check the returned metadata. Several entries may share a barcode; the Discogs id identifies the catalog entry itself.

Why are barcode or image fields null?

The source may not provide them. Some releases do not carry a barcode, and anonymous responses may omit images. The actor preserves null values rather than filling them with estimates or substitute URLs.

Does it return record prices or sales?

No. It returns database search results and collector counts. Marketplace listings, prices, transaction history and units sold are outside this actor's output.

Is this the official Discogs API?

No. It reads the public Discogs database search and turns the results into dataset rows, so you can use it like a Discogs API without writing code, managing tokens or handling pagination. Developers who need endpoints beyond search can use the official API alongside it.

Can I export Discogs data to CSV or Excel?

Yes. Every run writes to an Apify dataset, which you can download as JSON, CSV, Excel, XML or HTML, or read through the Apify API.

Why did I receive fewer rows than the limit?

A limit is a maximum. The search may have fewer matches, source pages may overlap, or a later page may fail. Read totalMatches, pagesFetched, samplingNote and any partial-result warnings in the run summary.