InvenioRDM: The Software Behind Zenodo

Most CASRAI comparisons involving Zenodo — Zenodo vs. Figshare, Zenodo vs. Dryad, institutional repository vs. Zenodo — answer a deposit question: where should a researcher send a dataset? This guide answers a different question, aimed at research offices and library IT rather than individual depositors: what software does Zenodo itself run on, and can an institution use that same software to operate its own repository instead of (or alongside) depositing into Zenodo? The answer is InvenioRDM, and a growing number of universities and research organizations now run their own instance of it.

What InvenioRDM is

InvenioRDM is a free, open-source repository application, released under the MIT license, that an institution installs and operates itself to run a research data management (RDM) repository. The Invenio project’s own site describes it as a “turn-key research data management repository” — meaning it ships as a working, deployable repository out of the box, with a modern web interface, rather than as a bare framework a developer has to build a user-facing site on top of. It is built on the underlying Invenio Framework, a toolbox of software components (record storage, search indexing, persistent identifiers, access control, and so on) that InvenioRDM assembles into a ready-to-run RDM platform. The same framework also powers InvenioILS, a separate integrated library system product, and can be used directly by developers building fully custom repository applications.

How InvenioRDM relates to Zenodo

Zenodo is not a competitor to InvenioRDM — it is InvenioRDM in production. Zenodo is a hosted service, operated by CERN with funding from the EU-backed OpenAIRE program, that anyone can deposit into without installing anything. InvenioRDM is the underlying application that Zenodo (along with CERN’s other repositories, including CDS Videos and CERN Open Data) is built on. The practical distinction: depositing in Zenodo puts a record on CERN’s shared, discipline-agnostic repository under CERN’s operational control; running an InvenioRDM instance means an institution installs and operates the same underlying software itself, under its own domain, branding, access policies, and retention decisions.

Who governs and funds it

InvenioRDM is developed as an open collaboration rather than a single vendor’s product. Core project partners include CERN — where the Invenio codebase originated — and Northwestern University, alongside a wider set of contributing research institutions. Development has been funded through a mix of the project partners’ own contributions and grants, including from CERN Knowledge Transfer, the European Union’s Horizon 2020 programme, the Alfred P. Sloan Foundation, and the Arcadia Fund. Because it is MIT-licensed and openly governed, no single institution can restrict who deploys it or how — the same property that lets CERN, a university library, and a small research center all run independent instances of the same underlying codebase.

Institutions running their own InvenioRDM repository

InvenioRDM is young relative to DSpace or EPrints — active institutional adoption dates mainly from the early-to-mid 2020s — but it already has real production deployments documented directly by the Invenio project, including:

CERN — repository.cern, CERN’s own general-purpose institutional repository (distinct from Zenodo, which CERN also operates and hosts).
Northwestern University — PRISM, Northwestern’s research data and scholarly output repository.
TU Graz — repository.tugraz.at, the Graz University of Technology’s institutional repository.
TU Wien — researchdata.tuwien.ac.at, Vienna University of Technology’s research data repository.

Each of these is a separately branded, separately administered repository — not a shared multi-tenant service — built on the same open-source codebase Zenodo uses. This is the core reason InvenioRDM is worth knowing about distinctly from Zenodo itself: it is the option available to an institution that wants Zenodo-like functionality (rich metadata, DOI minting, versioning, a modern deposit workflow) under its own institutional control, rather than depositing into someone else’s platform.

What an institution actually gets

Per the Invenio project’s own feature documentation, InvenioRDM provides, out of the box:

Persistent identifiers and metadata — automatic DOI minting through DataCite, plus integration with ORCID and OpenAIRE for author and funder identification and reporting.
Versioning and communities — records can be revised with a visible version history, and institutions can organize submissions into curated “communities” with their own review workflows, similar in concept to Zenodo’s own community feature.
Institutional integration — single sign-on (SSO) against an institution’s existing identity provider, and custom metadata fields an institution can add for local reporting needs.
Scale — the project describes the architecture as designed to run “from 1 to 100 million records,” i.e. suitable for anything from a small departmental archive to a large national-scale repository.
Pre-configured repository profiles — starting configurations tuned for different repository types (general institutional repository, data repository, records/documents repository) rather than a single fixed data model.

InvenioRDM vs. DSpace, EPrints, and Fedora

Institutions evaluating open-source repository software will inevitably weigh InvenioRDM against the incumbents covered in CASRAI’s own DSpace vs. EPrints vs. Digital Commons vs. Fedora comparison. The short version: DSpace (2002) and EPrints (2000) predate InvenioRDM by two decades and have far larger installed bases, more third-party hosting partners, and more institutional track record; InvenioRDM is newer, built specifically around research-data-management workflows (rich, DataCite-aligned metadata and dataset versioning) rather than the general document-and-thesis repository model DSpace and EPrints originated in, and it is the same codebase that already runs Zenodo at very large scale — a level of production battle-testing a brand-new platform would not have. Fedora, as covered in that comparison, is a different category entirely: a back-end object store that needs a separate application layer (Islandora or Hyrax) before it functions as a repository, whereas InvenioRDM is deployable on its own. None of this makes one platform categorically “better” — it is the same self-hosted-vs-vendor-hosted, technical-capacity-dependent trade-off research offices already navigate with DSpace and EPrints, just with a newer, RDM-focused option added to the shortlist.

What adopting InvenioRDM actually involves

Because InvenioRDM is free, open-source software rather than a subscription service, “adopting” it means installing, configuring, and operating it — the same technical-burden trade-off that applies to self-hosted DSpace or EPrints. An institution needs either in-house technical capacity to deploy and maintain the application, or a partner to do it on their behalf; unlike DSpace and EPrints, which have an established roster of long-running commercial hosting partners, InvenioRDM’s hosting-partner ecosystem is newer and smaller, reflecting the platform’s more recent adoption curve. Research offices and library IT teams evaluating it should budget for this as an operational commitment — deployment, upgrades, and metadata configuration — not a one-time purchase, in the same way they would for any self-hosted open-source repository platform.

Frequently asked questions

Is InvenioRDM free to use?

Yes. It is released under the MIT license, a permissive open-source license with no licensing fee. An institution still bears its own hosting and technical-support costs to run it, the same as with any self-hosted open-source repository platform.

What is the difference between InvenioRDM and Zenodo?

Zenodo is a specific hosted repository service, operated by CERN, that anyone can deposit into. InvenioRDM is the open-source software Zenodo is built on, which an institution can install and run independently to operate its own, separately branded repository rather than depositing into CERN’s shared one.

Who maintains InvenioRDM?

It is developed as an open collaboration whose core partners include CERN and Northwestern University, alongside other contributing institutions, with funding from sources including CERN Knowledge Transfer, the EU’s Horizon 2020 programme, the Alfred P. Sloan Foundation, and the Arcadia Fund.

Can an institution migrate from DSpace or EPrints to InvenioRDM?

Migration between repository platforms is technically possible but non-trivial — it involves mapping metadata schemas, re-minting or preserving persistent identifiers, and moving file content, which is true of any cross-platform repository migration, not something specific to InvenioRDM. Institutions considering it should treat it as a project on the same scale as any other repository-platform migration, and consult the Invenio project’s own documentation and community for current migration tooling and guidance.

Does an InvenioRDM repository mint DOIs?

Yes — InvenioRDM integrates with DataCite to mint DOIs for deposited records automatically, the same DOI-minting relationship Zenodo itself uses, since Zenodo runs on InvenioRDM.