Direct comparison
Dryad vs. Dataverse: Which Fits Your Data?
Dryad is a curated, fee-based, journal-linked repository. Dataverse is free, open-source software institutions self-host or use via Harvard Dataverse.
Side-by-side comparison
| Dimension | Dryad | Dataverse |
|---|---|---|
| What it is | One repository, operated by a single nonprofit organization | Open-source software; many independent installations run it, each its own repository |
| Operator | Dryad (nonprofit), partnered with the California Digital Library (CDL) | Developed by Harvard's Institute for Quantitative Social Science (IQSS); each installation is run by the institution that deploys it |
| Origin | Launched 2008/2009 (grew out of the Joint Data Archiving Policy, JDAP) | Coding began 2006 as the Dataverse Network Project, building on the Virtual Data Center project (1999-2006); rebuilt as the modern Dataverse 4 web application in 2015 |
| Software model | Closed platform — you use Dryad's hosted service, you don't run your own copy | Open source (Apache License 2.0) — self-host your own installation, or use an existing one such as Harvard Dataverse |
| Branding | Always “Dryad,” regardless of depositor's institution | Institution-branded — each installation (Harvard Dataverse, and many university- or consortium-run installations worldwide) carries its own name and collection structure |
| Curation model | Mandatory human curator review before every dataset publishes (file integrity, metadata completeness, reusability) | Configurable per collection — an installation's admins can require a “submit for review” approval step, or allow self-service publication with no review |
| Typical deposit trigger | A dataset tied to a specific manuscript, usually via direct integration with a journal's submission system | Varies by installation — individual researcher deposits, lab/project collections, or journal-linked collections, depending on how the hosting institution configures it |
| Cost | Data Publishing Charge on a sliding scale by size, unless sponsored by a partner institution, publisher, or society | Software itself is free; Harvard Dataverse is free to use (up to 1 TB per dataset). Self-hosting elsewhere costs the hosting institution's own infrastructure and staff time, not a per-deposit fee to the depositor |
| Example fee (unsponsored deposit) | $150 up to 5 GB; $520 up to 50 GB; $6,077 up to 1 TB (2025 schedule) | No per-deposit fee on Harvard Dataverse or most institutional instances; some installations may set their own policies |
| Default license | CC0 (public domain) by default | CC0 strongly encouraged as the default on Harvard Dataverse; individual installations and depositors can select other terms |
| Journal manuscript-system integration | Direct integration with Editorial Manager and eJournalPress, used by many journals as their designated data repository | No universal built-in manuscript-system integration; some journals run their own dedicated Dataverse collection instead |
| DOI registrar | DataCite | DataCite (each installation registers its own DOI prefix) |
| Preservation layer | Preservation copy held in Merritt (California Digital Library), which is CoreTrustSeal certified — Dryad itself is not separately certified, Merritt is | Depends on the installation — preservation guarantees are whatever the hosting institution commits to; several individual installations pursue their own certification |
| Best fit | A dataset backing a specific published paper, especially where the journal already integrates with Dryad | An institution, consortium, or journal that wants to run and brand its own repository, or a researcher whose institution already operates a Dataverse instance |
Common questions
FAQ
Is Dataverse free to use?+
The Dataverse software itself is free and open source (Apache License 2.0). Whether depositing into a specific installation is free to you depends on that installation's policy — Harvard Dataverse, the flagship instance, is free to use for datasets up to 1 TB. An institution that self-hosts its own installation bears the infrastructure and staff cost, not a per-deposit fee charged to depositors.
Can I use Dataverse if my institution doesn't run its own installation?+
Yes — Harvard Dataverse accepts deposits from researchers worldwide, not only Harvard-affiliated researchers, and is free to use.
Does Dataverse curate or review datasets before publication the way Dryad does?+
Not by default, and not uniformly. Dryad requires human curator review for every deposit. Dataverse's software supports an optional “submit for review” workflow that a collection's administrators can turn on, but plenty of installations and collections allow self-service publication with no review step at all — the level of curation depends on how the specific installation and collection are configured, not on the software itself.
Why would an institution choose to run its own Dataverse installation instead of just using Dryad or a generalist repository like Zenodo?+
Running an own-branded installation gives an institution control over its own collection structure, access policies, curation workflow, and integration with local systems — useful for a university, funder, or journal that wants a repository under its own name rather than depositing into someone else's platform. It's a different tradeoff than Dryad's model, where the curation and infrastructure work is centralized and paid for per deposit.
Do Dryad and Dataverse both mint DOIs?+
Yes — both register dataset DOIs through DataCite. Each Dataverse installation registers under its own DOI prefix.
