The public SPARQL endpoint
https://sparql.plugin-universe.com/public/query
Read-only, open, and CC0. Ask it anything the catalogue can answer.
SELECT ?name ?vendor WHERE {
?plugin a <http://purl.org/stuff/transmissions/PluginProfile> ;
<http://www.w3.org/2000/01/rdf-schema#label> ?name .
OPTIONAL { ?plugin <http://purl.org/stuff/transmissions/vendor> ?vendor }
} LIMIT 20
The default graph is the union of everything here
A query with no GRAPH clause sees every named graph at once, so
SELECT * WHERE { ?s ?p ?o } does what you would expect. This is worth stating
because the alternative — an empty default graph — is a common configuration and
makes a newcomer's first query return nothing, which reads as a broken endpoint
rather than as a setting.
Provenance is kept, so a query that cares which source a statement came from can name its graph:
SELECT ?g (COUNT(*) AS ?triples) WHERE { GRAPH ?g { ?s ?p ?o } } GROUP BY ?g
What is here, and what is not
This is a published copy, rebuilt from the catalogue rather than being the catalogue. It holds the graphs whose licence permits redistribution, in three parts: facts under CC0, material from permissively-licensed sources with their notices, and contributor-written prose under CC BY-SA.
Accounts and pending contributions are not here at all. They are not filtered out on the way to you; they were never loaded into this dataset. That is deliberate — a filter is something somebody has to maintain, and absence is not.
Because it is a copy, it lags the live catalogue by up to a day.
Limits, and the better path for bulk
Queries time out at thirty seconds and the endpoint is rate-limited. Both exist because an unbounded query costs this server far more than it costs whoever sent it, and this runs on one small machine.
If you want the whole dataset, take the dump instead of paginating through this. It is faster for you and cheaper for us, and it is the same data:
- the dumps, one file per graph, with a manifest and a VoID description
- an Open Audio Stack compatible registry, if you are a package manager
Terms
Facts are CC0 — public domain, attribution requested but not required. Contributor prose is CC BY-SA 4.0 and must be attributed. The parts are separate datasets so the boundary is a property of the data rather than something you have to work out.
Credit, if you would like to give it: Plugin Universe, https://plugin-universe.com