What Zenodo is
Zenodo is a general-purpose research repository operated by CERN and funded through European open-science programmes. You upload a research output of almost any kind — a dataset, a software release, a poster, a slide deck, a protocol, a set of negative results — and it issues a DOI and preserves the record.
The two properties that make it the default recommendation are its steward and its scope. CERN is an intergovernmental organisation with a preservation commitment measured in decades and no commercial pressure to pivot, which is a materially different guarantee from a company’s terms of service. And Zenodo accepts anything: outputs that no journal will publish and no disciplinary repository will take still get a persistent identifier, which is how the unglamorous 80% of research work becomes citable at all.
Why researchers use it
- A DOI for things that had none — the protocol, the poster, the analysis script, the dataset your paper depends on, all now citable objects with your name on them.
- Automatic GitHub archiving — flip a switch on a repository and every tagged release is archived with its own DOI, which turns “code available on GitHub” into a citable, permanent artefact.
- Concept DOIs — one identifier always resolving to the latest version, plus per-version DOIs for exact reproducibility statements. Cite the first in a paper, the second in a methods appendix.
- Generous size limits — tens of gigabytes per record by default, with larger allowances on request.
- It satisfies funder mandates — a data-management plan requirement met in an afternoon rather than a negotiation.
Where it fits in a research workflow
Zenodo is the archival endpoint. Work happens in GitHub, on your analysis machine, and in your institutional storage; Zenodo is where the version of record goes when the paper is accepted.
The pattern worth setting up once: link your repository to Zenodo, tag a release at submission, and cite the resulting DOI in the data availability statement. Live code lives on GitHub, where it will change; the archived snapshot lives on Zenodo, where it will not — and a reviewer in three years needs the second. Large data tracked with DVC during the project gets deposited here at the end. Where it hands off: OSF if you also need preregistration and project structure, Figshare if your institution’s portal feeds university reporting, and a disciplinary repository — GenBank, ICPSR, PANGAEA — if one exists for your data type.
Getting started
Half an hour, and set up the GitHub link before you need it.
- Sign in with ORCID so your outputs attach to your identifier automatically. Doing this at the start saves reconciling records later.
- Turn on the GitHub integration for your analysis repository, then create a release. The DOI appears without further action, and this is the single highest-value ten minutes in the whole tool.
- Upload with complete metadata — all authors with ORCIDs, a deliberate licence, grant numbers, and a description someone outside your lab can parse. Nobody curates this for you.
- The step people skip: reserve the DOI before submission. Zenodo can issue one ahead of publishing the record, which means you can cite your own dataset in the manuscript rather than adding “DOI to follow” at proof stage.
Zenodo vs the alternatives
| Alternative | Does it better | Pick it if |
|---|---|---|
| Figshare | Institutional portals, usage metrics, and a slicker interface | Your university runs a portal that feeds its reporting |
| OSF | Preregistration, project structure, and research-native organisation | The output is a study, not a file |
| A disciplinary repository | Curation, community search, and domain metadata standards | One exists for your data type — use it first |
| GitHub alone | Nothing archival — repositories can be deleted | You have not yet reached publication |
Zenodo is the universal fallback and the right default. A curated disciplinary repository beats it whenever one exists.
Cost, licensing, and your data
Free, with no paid tier, funded through CERN and European open-science programmes rather than by users. Deposits are covered by a stated preservation policy tied to CERN’s own data-centre lifetime, which is the strongest continuity commitment available to an individual researcher at no cost.
Licensing is your decision and worth making deliberately rather than accepting a default. CC-BY is conventional for data and is a real choice with consequences for reuse and attribution; restrictive licences on publicly funded data may conflict with your funder’s terms. The harder question is what you are permitted to deposit at all: participant-derived data usually cannot go into an open repository without appropriate consent and de-identification, and Zenodo’s closed and restricted access options exist precisely so that a record can be citable without the underlying files being public. Use them rather than deciding not to deposit.
The honest review
Strengths. Zenodo made every research output citable, for free, on infrastructure with a credible multi-decade horizon. The GitHub integration in particular converts a routine act — tagging a release — into a permanent, citable artefact with no further effort, which is the rare case of good practice becoming the path of least resistance.
Limitations. Discovery is genuinely weak. Zenodo’s search is not where people look for datasets, so a record’s visibility comes almost entirely from the links you put in your paper — deposit and expect nothing else. There is no curation, which means no quality signal and no metadata review: a well-documented dataset and an unlabelled zip file sit side by side. Metadata quality is entirely on you, and most deposits are poor. And it is a generalist, so it lacks the domain-specific metadata standards that make disciplinary repositories genuinely searchable.
Verdict. Adopt it for every paper you publish — it costs an afternoon once and makes your outputs citable permanently. Skip it as your primary deposit only when a curated disciplinary repository covers your data type, in which case deposit there first and optionally mirror here. The condition that flips the answer is whether your field has its own infrastructure.
Common questions
Is Zenodo free?
Yes, entirely, with no paid tier. It is operated by CERN and funded through European open-science programmes, with generous per-record size limits and larger allowances available on request.
How do I get a DOI for my GitHub repository?
Enable the Zenodo integration for the repository, then create a tagged release on GitHub. Zenodo archives that snapshot automatically and mints a DOI. You get both a concept DOI that always points to the latest version and a version-specific DOI for exact citation.
Zenodo or Figshare?
Zenodo for its non-commercial steward and stronger long-term guarantee. Figshare when your institution runs a portal that feeds university reporting, which makes it the path of least resistance. If a curated disciplinary repository exists for your data type, that beats both.