QLever
activeby University of Freiburg (Chair of Algorithms and Data Structures) · since 2017
Fast open-source SPARQL engine that serves all of Wikidata or OpenStreetMap from one machine.
Overview
QLever is an open-source SPARQL engine from the University of Freiburg's algorithms group, written in C++ and designed to load and query very large RDF datasets (full Wikidata, UniProt, OpenStreetMap, PubChem) on a single commodity server. It builds compressed, sorted index permutations ahead of time and adds text search over literals and GeoSPARQL-style spatial queries, often answering complex queries in well under a second where older triplestores time out. SPARQL Update support has landed recently. It is the leading candidate for heavy read-mostly public endpoints, but it is still pre-1.0 and has no clustering.
Strengths & trade-offs
Strengths
- +Loads and queries full Wikidata-scale datasets on a single commodity machine
- +Complex SPARQL queries often finish in sub-second time where other engines time out
- +Built-in text search and spatial predicates over OpenStreetMap-style data
- +Simple qlever CLI automates download, indexing and serving
Trade-offs
- −Pre-1.0: SPARQL Update and some features are still maturing
- −Index builds are offline and take hours for the largest datasets
- −No clustering, replication or access control built in
Great for
- +Public SPARQL endpoints for Wikidata, OSM, UniProt
- +Large read-mostly knowledge graphs
- +Research on big RDF datasets
- +Replacing aging Blazegraph deployments
Think twice for
- −Write-heavy transactional workloads
- −Property-graph apps
- −Managed-service users
Capabilities
Native or official first-party support · 4 of 25 tracked
Scale & consistency
- Comfortable data size
- Large · ≤ 1 PB
- Horizontal scaling
- None
- Multi-region
- No
- Transactions
- Limited ACID
- Consistency
- Strong
- Schema
- Schemaless
- Architecture
- Single node, Client server
- Query languages
- SPARQL
Tens of billions of triples (full Wikidata and larger) on one machine with fast SSDs.
Deployment
Ecosystem
Alternatives to QLever
Hybrid SQL/RDF engine that runs many of the largest public SPARQL endpoints.
qlever vs virtuoso →Abandoned Java RDF/SPARQL store, still known as the engine behind Wikidata Query Service.
qlever vs blazegraph →Apache Jena's native RDF store, embeddable in Java or served over SPARQL with Fuseki.
qlever vs apache-jena-tdb →Small, fast SPARQL database in Rust that embeds in Rust, Python and JavaScript or runs as a server.
qlever vs oxigraph →Standards-compliant RDF triplestore with reasoning, connectors and MCP support for GraphRAG.
qlever vs ontotext-graphdb →Eclipse Foundation's Java RDF framework with native, LMDB and in-memory SPARQL stores.
qlever vs eclipse-rdf4j →QLever FAQ
What is QLever?
QLever is a RDF / triple store database by University of Freiburg (Chair of Algorithms and Data Structures), first released in 2017. QLever is an open-source SPARQL engine from the University of Freiburg's algorithms group, written in C++ and designed to load and query very large RDF datasets (full Wikidata, UniProt, OpenStreetMap, PubChem) on a single commodity server. It builds compressed, sorted index permutations ahead of time and adds text search over literals and GeoSPARQL-style spatial queries, often answering complex queries in well under a second where older triplestores time out. SPARQL Update support has landed recently. It is the leading candidate for heavy read-mostly public endpoints, but it is still pre-1.0 and has no clustering.
Is QLever open source?
Yes. QLever is open source. License: Apache License 2.0.
Does QLever support ACID transactions?
Partly. QLever offers limited ACID guarantees (for example, single-record or single-partition transactions), with strong consistency.
Does QLever scale horizontally?
QLever does not scale horizontally; it runs on a single node, so you scale it up rather than out. It is comfortable at large scale (≤ 1 PB).
How can I run QLever?
You can self-host it.
What is QLever best for?
QLever is a strong fit for public SPARQL endpoints for Wikidata, OSM, UniProt; large read-mostly knowledge graphs; and research on big RDF datasets. Think twice for write-heavy transactional workloads.
What are the best alternatives to QLever?
Common alternatives to QLever include OpenLink Virtuoso, Blazegraph, Apache Jena TDB, Oxigraph and Graphwise GraphDB (Ontotext). The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.
How popular is QLever?
QLever ranks #186 of 844 databases on the WhichDB Score (17.9/100), and #3 among RDF / triple store databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.