Open research data
Data that outlives the grant that funded it
Up to 20 TB per deposit, a DOI on publication, three geographically separate mirrors, and no registration wall between a reader and the files.
What a repository owes its depositors
A citable identifier
A DataCite DOI at publication, versioned so a corrected v2 does not orphan the citation of v1.
Preservation, in writing
Fixity checked monthly, three copies on two continents, and a funded commitment to keep everything for at least 25 years.
Open by default
No account, no click-through, no rate limit that punishes a whole-dataset download. Embargo is available and time-boxed.
Big files welcome
20 TB per dataset and no per-file ceiling. Uploads resume, checksums are verified server-side.
Findable
Rich metadata, OAI-PMH, a public search API and an index that other aggregators harvest hourly.
Linked to the paper
Bidirectional links to publications, software and grants, so the dataset is not an orphan.
Cost
Who pays for nine petabytes
We are funded by institutional membership and a national research infrastructure grant. Depositors pay nothing and readers pay nothing, which is the only arrangement that makes an archive actually get used.
- 61 member institutions, annual fee scaled to size
- A national infrastructure grant covering storage growth
- No article-processing-style charge, ever
- Costs published annually in the transparency report