# Virtuoso Open Source puts relational, graph and document data in one engine > The Virtuoso Open Source Edition is a database server written in C that puts three data models inside one engine, relational, graph and document, and builds three more products on top of that core: a data integration layer with ODBC and JDBC, an HTTP application server, and a linked data platform with RDF and SPARQL. It is a mature codebase with a commercial sibling, and that shapes how you should read its documentation. **openlink/virtuoso-opensource** — Virtuoso is a high-performance and scalable Multi-Model RDBMS, Data Integration Middleware, Linked Data Deployment, and HTTP Application Server Platform - Repository: https://github.com/openlink/virtuoso-opensource - Website: https://vos.openlinksw.com - Stars: 976 · Forks: 228 - Language: C - License: NOASSERTION - Published: 2026-10-20 · Updated: 2026-10-20 · Language: en - Canonical page: https://hysenlabs.com/projects/openlink-virtuoso-opensource ## One engine, three data models Virtuoso Open Source Edition, abbreviated VOS, is described as a multi-model RDBMS platform providing high-performance and scalable data management, and its central claim is that it integrates relational, graph and document data models within a single database engine. That is the part that is hard to do and the reason the product exists. A relational engine stores tables, a graph engine stores triples and a document engine stores flexible JSON-like objects, and most products make you choose or federate. Virtuoso instead puts all three in one server so that a query can span them, which is the reason a linked data deployment is a configuration choice rather than a migration. The engine is written in C, which the repository records as the primary language, and the code is a large autotools project with the conventional autogen.sh and configure.ac layout and separate source directories for the applications, the libraries and the binaries. ## Data integration through ODBC and JDBC The second product built on the engine is data integration middleware. VOS includes Data-Integration Middleware with support for ODBC and JDBC, which facilitates connection and interaction between data sources, and it provides data exchange and transformation capabilities. That matters because it is the difference between a database you query and a database that can absorb the systems around it. A relational engine with an ODBC driver can present tables that are not stored in it, and a JDBC driver makes the same possible for a JVM application, so a migration or a federation can happen without changing the client side. The optional components section points at JDBC drivers and providers for Jena and RDF4j published on Maven Central, which is the detail that shows the linked data side is meant to be used from Java code as well, and there is a separate integration guide for GeoSPARQL, the geographic extension of SPARQL. ## An HTTP server with a semantic web layer The third and fourth products are the web layers. VOS includes an HTTP(S) Application Server Platform for deploying web applications and services, with built-in support for web standards and protocols. Building on that, it is a Linked Data deployment platform with support for RDF, SPARQL and semantic web applications. The layering is the interesting architecture, because the HTTP server is not a separate product bolted onto a database, it is the same process serving both, which is why a SPARQL endpoint can be a few lines of configuration rather than a separate service to deploy and secure. The documentation list includes a Database Administration Manual hosted on the OpenLink docs site, the VOS wiki, and implementation notes for a feature called SPARQL ANYTIME, which is a query mode that returns partial results with a time bound, useful for queries over large graphs where a full answer is not what you want. ## Four installation routes, three of them binaries The installation section offers a binary download for each desktop platform, with a Linux tarball for the generic glibc build, a macOS disk image and a Windows disk image, all linked from the GitHub releases page. On top of that there is a Docker route, described in a separate README file, a build from source with instructions for Linux and macOS in another README and a Windows build guide in yet another, and an upgrade document for moving from previous versions. The split into many README files is the documentation style of an old project, and it is a genuine friction point for a newcomer, because there is no single page that tells you which of these applies to you. The linked data driver packages are distributed through Maven Central rather than alongside the server, which is a sensible choice since the server itself is not built on Maven. ## The C codebase and how it is built The repository is a large C project with an autotools build. At the top level there is configure.ac, autogen.sh for the bootstrap, Makefile.am, and separate source trees in appsrc/, libsrc/ and binsrc/, with a windows/ directory for the platform specific build and a debian/ directory for packaging. The remaining odds and ends in the tree are the kind of file a mature project accumulates, an.indent.pro with the indentation rules, a.profile, a.cvsrc, plus AUTHORS, CREDITS, ChangeLog, INSTALL, NEWS.md and a long list of README files, each covering one topic such as the build, Docker, Windows, upgrades, GeoSPARQL, Jena, Hibernate, PHP 5 and the git repository policy. Two of those are worth a second look for a reader, since README.hibernate.md and README.jena.md tell you which Java and Java RDF stacks the project expects to interoperate with, which is more informative about the audience than any feature list. ## Releases, support and the open source and commercial split The project is clearly alive and the release cadence is steady, with v7.2.17 published on 2026-05-07, v7.2.16.1 on 2025-10-15 and v7.2.16 on 2025-10-14, and the last push on 2026-10-12. The default branch is develop/7, which follows the convention of a development branch per major version. The repository carries a SECURITY.md for reporting vulnerabilities, an issue tracker where reports should include a summary, a description and steps to reproduce, community forums, and a support case process that is available if you are covered by a support contract. The See Also section points at the commercial Virtuoso edition and at the OpenLink corporate site, which is the honest framing for an open source edition that is also the community tier of a commercial product. One licensing point needs care. The README states that VOS is licensed under the GNU General Public License version 2 dated June 1991, pointing at the COPYING file, while the project listing records a custom licence that GitHub cannot classify, so the COPYING file in the repository is the document to read, and the credits file is where the code incorporated from other projects is acknowledged. ## Conclusion Virtuoso Open Source Edition fits a team that has to serve SPARQL and relational queries over the same data, or that wants an application server with a database behind it in one install, and the ODBC and JDBC drivers are what make it reachable from an existing stack rather than an island. It does not fit a team that wants a modern package manager or a container-only workflow, since the build is autotools based and the Docker path is one file among several, and it does not fit anyone who needs a feature that only the commercial edition carries. Before you deploy it, read the upgrade notes if you are moving from an older release, check the SPARQL ANYTIME implementation notes if that query mode matters to you, and note the licence discrepancy, where the README states GPL version 2 while the project listing records a custom licence, so the COPYING file is the document to read. ## FAQ ### What is Virtuoso Open Source Edition? It is a multi-model RDBMS platform in C that integrates relational, graph and document data models within a single database engine, and it also includes data integration middleware with ODBC and JDBC support, an HTTP application server platform, and linked data deployment with RDF and SPARQL. ### How do I install Virtuoso? By downloading the installer for Linux, macOS or Windows from the GitHub releases page, by running it in Docker, or by building from source, with separate instructions for Linux and macOS and a Windows build guide. There is also an upgrade document for previous versions, and the JDBC drivers for Jena and RDF4j come from Maven Central. ### What is SPARQL ANYTIME? It is a query mode that the project documents in its own implementation notes, and it is a separate entry in the documentation list alongside the database administration manual and the VOS wiki, which suggests it is a feature with enough subtlety to need its own explanation. ### Is Virtuoso still maintained? The repository is not archived and the last push was on 2026-10-12, with releases v7.2.17 from 2026-05-07, v7.2.16.1 from 2025-10-15 and v7.2.16 from 2025-10-14, on the develop/7 branch. Support cases are available to customers with a support contract. ### What licence is Virtuoso Open Source Edition under? The README states the GNU General Public License version 2 dated June 1991 and points at the COPYING file, while the project listing records a custom licence that GitHub cannot classify, so the COPYING file in the repository is the document to read. The credits file records code incorporated from other projects. ## Sources - [Issues](https://github.com/openlink/virtuoso-opensource/issues) - [openlink/virtuoso-opensource on GitHub](https://github.com/openlink/virtuoso-opensource) - [Project website](https://vos.openlinksw.com) - [README](https://github.com/openlink/virtuoso-opensource/blob/develop/7/README.md) - [Releases](https://github.com/openlink/virtuoso-opensource/releases) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/openlink-virtuoso-opensource