Zenodo

Zenodo is a general-purpose open repository built and operated by CERN, created so that any researcher in any discipline can deposit data, software, figures, posters, presentations or reports and receive a citable DOI in return. It exists to catch everything that has no obvious home — the outputs that fall outside a domain repository and would otherwise live on a lab hard drive until the postdoc who made them leaves.

What it does

You upload a file, describe it, choose a licence, and Zenodo mints a DOI. That DOI is permanent and resolvable, and it is versioned: subsequent releases of the same dataset or software get their own DOI while a concept DOI continues to point at the latest version. Records can be embargoed until a paper appears, restricted to named parties, or made fully open, and communities let a lab, project or conference curate its own collection.

The GitHub integration is the feature most researchers meet first. Connect a repository, tag a release, and Zenodo archives that release and issues a DOI automatically — which is how a piece of analysis code becomes a citable object rather than a URL that will rot.

Why it matters

Funder and journal data-availability requirements have made deposit non-optional, and Zenodo is the path of least resistance for the majority of cases. It is backed by CERN’s infrastructure rather than a startup runway, files are stored with the same institutional commitment as physics data, and there is no charge for reasonable deposit sizes.

It is also the reference implementation of what a repository should do for reproducibility. A methods section that cites a versioned Zenodo DOI for its analysis code is unambiguous in a way that “code available on request” never is.

Strengths

• Free, with no publication or deposit charges for normal use.
• CERN-operated, which is about as durable as research infrastructure gets.
• DOI versioning, so citations survive updates to the underlying object.
• Automatic GitHub release archiving.
• Discipline-agnostic, accepting data, software, slides, posters and negative results alike.
• Communities allow curated collections without running your own repository.

Limitations

Being general-purpose is also the weakness: a domain repository will almost always offer richer, standardised metadata and better discoverability for its own data type. Sequence data belongs in the ENA or SRA, structures in the PDB, earth-science data in PANGAEA. Zenodo is where output goes when nothing more specific fits, or when you need a DOI quickly. Discovery within Zenodo is functional rather than good, and per-record file-size limits mean genuinely large datasets need a conversation rather than a drag-and-drop.

Verdict

Essential infrastructure, and one of the few tools on this site we would call a default rather than a choice. Connect your GitHub account to it today; the first time a reviewer asks for your analysis code you will already have a DOI to give them.