Back to all databases

MarkLogic

A multi-model NoSQL database with native RDF triple storage, SPARQL support, and semantic reasoning. Combines JSON, XML, and RDF data models with ACID transactions and integrated full-text search.

Overall rank
#52
First released
2001
License
Proprietary
Built with
C++

Details

Vendor
Progress Software
Type
Multiple
Category
Enterprise
License
Proprietary
Language
C++
Query Languages
SPARQLXQuerySQL
Protocols
HTTP/RESTSPARQL ProtocolPostgreSQL wireJDBC/ODBCNative binary
Sources 9 references

Protocols HTTP/REST, SPARQL Protocol, JDBC/ODBC, PostgreSQL wire, Native binary

MarkLogic ships five distinct client channels, confirmed independently: the generic REST Client API (document CRUD/search); a dedicated W3C SPARQL Protocol endpoint at /v1/graphs/sparql that is separate from the general REST surface; official ODBC and JDBC drivers whose app server (port 5432) literally speaks the PostgreSQL wire protocol; and the still-current XDBC/XCC proprietary wire protocol for Java middleware (confirmed present in the current v12 admin docs, not just legacy v9). pass_A missed SPARQL Protocol and JDBC; pass_B missed PostgreSQL wire and Native binary.

License Proprietary

Confirmed; no change from current value. MarkLogic is distributed as binary executable software under a commercial license, with no source code disclosure.

  • Licensor grants to You a limited, non-transferable, non-exclusive, internal use license to install, access, start and use Licensor software in binary executable form
    EULA - MarkLogic Community · reviewed 2026-08-02

Released 2001

Confirmed; no change from current value. MarkLogic's first product, Cerisent XQE 1, shipped in 2001 under the company's original name, Cerisent.

  • 2001 – Cerisent XQE 1: ACID transactions, Full-text search, XML Storage, XQuery, Role-based security
    MarkLogic Server · reviewed 2026-08-02

Language C++

Confirmed; no change from current value. MarkLogic's own technical documentation attributes its performance to the core engine's native C++ code; Wikipedia's infobox additionally lists C and JavaScript, but JavaScript there refers to the embedded V8 engine used for user application code, not the storage/query kernel.

  • Speed and scale are an obsession for MarkLogic. They're not features you can add after the fact; they have to be part of the core design. And they are, from the highly optimized native C++ code to the algorithms we'll discuss later.
    Inside MarkLogic Server · reviewed 2026-08-02