What is Koha? A complete guide to the open-source library system

Where Koha came from: one library's real problem

Koha was born from a concrete problem, not from a business plan. The Horowhenua Library Trust (HLT) in New Zealand was running an integrated library system more than twelve years old that no longer received any development. It was not ready for the year 2000 date change (the Y2K problem), it no longer met the library's needs, and buying a new commercial system required an investment the trust could not afford.

The answer was to build its own. HLT commissioned Katipo Communications, a New Zealand company, which developed the system during the last quarter of 1999. It went into production on January 1, 2000. HLT and Katipo then made the decision that defined the project's future: they released the code under the GNU General Public License (GPL), so that any other library could use it, modify it and contribute to its development. They named it Koha, a Māori word meaning "gift" or "donation", in direct reference to that decision. In its first year the project received the 3M Award for Innovation in Libraries and an award in the community and non-profit category of the ANZ Interactive Awards.

From a New Zealand project to a global community

Development did not stay in New Zealand. From 2000 onward, companies around the world began offering commercial support for Koha; today more than fifty companies do so in different countries, and BiblioLabTech is one of them. A decisive contribution arrived in 2001, when French developer Paul Poulain began adding substantial features. The most important was multilingual support: by 2010, Koha had been translated from the original English into French, Chinese, Arabic and several other languages. Support for the MARC bibliographic standards and the Z39.50 search protocol came in 2002, a key step that let Koha interoperate with the rest of the world's library ecosystem.

Today Koha has no owner company. It is governed by an international community of libraries of every type and size, volunteer developers and the companies that provide commercial support, coordinated through koha-community.org. Anyone can propose changes, report bugs or contribute code, and decisions about the project's direction are taken in open community forums, not in one company's boardroom. The community publishes two feature releases a year, in May and November, plus monthly maintenance releases.

That is the most important structural difference from a proprietary ILS. It is also why no provider, including any that offers "its own version" of Koha with proprietary add-ons, can charge for the software itself, only for the services around it. The community's paid-support directory is explicit: any organization that claims to be an "official" Koha support provider is misrepresenting its relationship with the project.

What is Koha, and what is it not?

• Koha is an integrated library system (ILS). It centralizes cataloging, circulation, acquisitions, serials and the public catalog in one web platform. For the broader picture of this software category, read our guide to the integrated library system.
• Koha is not an institutional repository like DSpace. It is not designed to preserve theses, articles or research data in open access, or to be harvested by academic aggregators. If your institution needs both a library catalog and an academic repository, the normal setup is to run both systems in parallel; see what DSpace is for that other piece.
• Koha does not depend on a single manufacturer. Because it is open source under the GPL, any provider can implement, customize or support it, which avoids the vendor lock-in typical of proprietary software.

How the Koha library system is built

Koha is a web application written mainly in Perl, with a relational database in MySQL or MariaDB; MariaDB is the recommended option for modern installations. Bibliographic records are stored in MARC, MARC21 by default with UNIMARC also supported, and they are accessible through the standard Z39.50 and SRU protocols, so catalogs can be imported and exported between libraries in any country.

Search and indexing have evolved over time. The original engine, still the default, is Zebra. The project later added Elasticsearch, and more recently OpenSearch, as an alternative indexing engine for larger installations or for libraries that need more flexible searching. The choice is made in the system configuration. A modern installation also includes performance components such as Memcached and Plack, and recent versions use a message broker, RabbitMQ, to handle background jobs such as bulk notice sending.

This stack (Perl, MariaDB, Zebra or Elasticsearch) is lighter than that of a repository such as DSpace (Java, PostgreSQL, Solr and Angular), which in practice means more modest server requirements. That does not mean running Koha in production takes no expertise. Index configuration, upgrade management and template customization still demand specialized knowledge, which is why many libraries choose managed support over running Koha with their own staff. We compare both options in our guide to Koha hosting.

Koha modules

Koha brings together in one system the functions that many small libraries still run separately or in spreadsheets:

A distinctive trait of Koha compared with many proprietary systems is that these modules are not separate products sold one by one: they come integrated in a single codebase from the first installation. That simplifies maintenance, because there are no updates to synchronize between modules bought at different times. It also means that enabling or disabling functions depends on system configuration, not on negotiating with a vendor which package to buy.

Beyond the core modules, Koha includes functions many libraries never use simply because they do not know they exist: holds between branches in library consortia, automatic email and SMS notices, and standard interfaces to connect with other systems, such as SIP2 for self-check machines, OAI-PMH for metadata harvesting, and ILS-DI and a REST API for integration with academic or identity systems.

Who uses Koha, and at what scale

Adoption figures for Koha vary by source. There is no single, centralized official count at koha-community.org, but every source points to broad and sustained adoption since 2000. Specialized studies and directories such as Library Technology Guides document several thousand registered installations worldwide, and commercial providers estimate higher numbers when installations that were never formally reported are included.

What is well documented are specific large-scale deployments. In 2014, Turkey adopted an adapted version of Koha to automate 1,136 public libraries, managing more than 17 million items and serving about 2 million users: one of the largest documented open-source ILS deployments anywhere. That case is a reminder that Koha is neither small nor experimental. Sustaining 17 million items and 2 million active users requires an indexing engine and a database that can handle real load, which is the same technical reason the project added Elasticsearch as an alternative to Zebra over the years.

By type of institution, Koha runs in public, school and special libraries and in academic libraries of every size, including large universities. The specialist literature describes it as the leading open-source ILS particularly in the developing world, including several national library automation initiatives. In Latin America, where BiblioLabTech is based, Koha is now the most established open-source alternative to the proprietary systems that dominated the market a decade ago.

Why the project did not disappear

The history of free software is full of promising projects that stopped being maintained when their original creator lost interest or funding. Koha avoided that fate through a combination of factors worth naming:

1. A GPL license from day one. The trust's decision to license the code under the GPL, instead of keeping a closed custom build, allowed other institutions and companies to invest in the project without depending on HLT.
2. Contributors from outside New Zealand. The arrival of external contributors with real technical capacity, such as Paul Poulain in France, proved the project could grow beyond its country of origin.
3. A commercial support market. Companies that could charge for installing, migrating and maintaining Koha without owning the code gave the project a sustainable economic model without selling licenses, an idea that was still relatively new for library software in 2000.

How Koha compares with other library systems

The real decision a library faces is not "Koha, yes or no" in the abstract, but Koha against a handful of concrete alternatives, each with a different model:

Koha's advantage over other open-source ILSs is not only technical; it is its ecosystem. It has one of the largest development communities in its category, a wide network of commercial support companies in many countries and continuous development since 2000. That reduces the risk of abandonment, which is a real risk in niche software with small communities.

What you actually contract when you adopt Koha

As with any open-source software, no company sells Koha licenses, because there are none: the software is a free download. What a library contracts is the service around it: installation and initial configuration, migration of the catalog from the previous system, customization of templates and circulation rules, hosting in the cloud or on its own server, and ongoing support for incidents and version upgrades. That is why the market for Koha providers looks more like a professional-services market than a product market.

BiblioLabTech implements Koha for academic, school and public libraries, including migrations from proprietary systems and upgrades of old versions without losing historical data. Every Koha project by BiblioLabTech includes BiblioScan AI, our AI-assisted MARC21 cataloging tool, at no extra cost. The Escuela Colombiana de Ingeniería Julio Garavito runs its catalog on Koha with BiblioScan AI cataloging, on a dedicated server administered by our team; read the case study.

If your library already has a system and is weighing a move, start with how to migrate to Koha. If you are sizing a budget, the cost structure is in Koha implementation cost, and the project plan is in our Koha implementation guide. If you are still unsure whether Koha fits your case, given your collection size, number of branches or the integrations you need, our Koha service starts with an initial assessment at no cost. We work remotely with libraries in any country, in English, Spanish, Portuguese and French.