Best Preprint Servers and Open Repositories 2026: Where to Share Your Research

You have finished a study, and the journal review process stretches months ahead — yet your findings are ready now. The best preprint servers and repositories let you post your work immediately, get a citable DOI, and receive community feedback while the formal peer-review clock ticks down. The question is not whether to use them but which one fits your discipline, your data type, and the policies of the journal you plan to submit to. This directory maps every major option for 2026.

Preprint servers host manuscript drafts — the full paper before (or alongside) peer review. General repositories store datasets, code, figures, and other research outputs. Many researchers use both: a preprint server for the article and a repository for the underlying data. Understanding the distinction helps you choose the right platform for each output and stay compliant with funder mandates, which increasingly require immediate open sharing of both manuscripts and data.

Preprint Servers: Discipline-by-Discipline

A preprint server accepts manuscript files, applies basic screening (usually for scope and obvious integrity violations, not peer review), assigns a DOI, and makes the paper publicly available — typically within one to three business days. Most major preprint servers are indexed by Google Scholar and CrossRef, so your preprint becomes citable and discoverable from day one.

Journals’ policies on preprints vary. The overwhelming majority of journals in physics, biology, medicine, and social sciences now accept submissions of manuscripts that have been posted as preprints. The SHERPA/RoMEO database lets you check a specific journal’s preprint policy before posting.

1. arXiv

URL: arxiv.org
Disciplines: Physics, mathematics, computer science, quantitative biology, quantitative finance, statistics, electrical engineering, economics
Operator: Cornell University (non-profit)
Cost: Free to submit; free to read

arXiv is the oldest and most established preprint server in the world, launched in 1991 for physics and now home to more than two million papers. In many fields — particularly high-energy physics, machine learning, and pure mathematics — posting to arXiv is a standard step in the publication workflow, not an optional extra. Papers appear in subject-specific feeds that thousands of researchers check daily.

Screening: arXiv does not peer-review papers, but moderators check for scientific relevance and scope. New submitters need endorsement from an established arXiv author in the relevant subject area. As of February 2026, all submissions must include a full English-language version (original or translation).

Licensing: arXiv offers six licence options. Authors retain copyright under all but CC Zero. The default is the arXiv Perpetual Non-Exclusive License, which limits reuse to arXiv’s own distribution. For maximum openness, choose CC BY 4.0 (allows reuse with attribution, including commercial) or CC BY-SA 4.0 (share-alike). The most restrictive Creative Commons option available is CC BY-NC-ND 4.0. Check your publisher’s policy before choosing: some publishers require a specific licence for accepted manuscripts. Source: arXiv licence information.

Indexing: Google Scholar, CrossRef, NASA ADS, Semantic Scholar, and field-specific aggregators. arXiv papers are not indexed in PubMed or Scopus by default.

Best for: Researchers in STEM who want maximum visibility before journal submission, especially in physics, CS, and maths where arXiv is the de facto communication channel.

2. bioRxiv

URL: biorxiv.org
Disciplines: All of biology — genetics, neuroscience, microbiology, cell biology, ecology, bioinformatics, plant science, and more
Operator: openRxiv (non-profit; took over stewardship from Cold Spring Harbor Laboratory in 2025/2026)
Cost: Free

bioRxiv launched in 2013 and rapidly became the default preprint server for the life sciences. Papers are screened for scope and basic research integrity but are not peer-reviewed. Screening typically takes one to two business days. Authors can post new versions of a paper at any time; all versions remain publicly available with version tracking.

Licensing: Authors choose from CC BY, CC BY-NC, CC BY-ND, CC BY-NC-ND, or CC BY-NC-SA. The recommended default for maximum openness is CC BY 4.0.

Indexing: Google Scholar, CrossRef, Semantic Scholar, Europe PMC. bioRxiv papers are not indexed in PubMed directly, but Europe PMC links preprints to their published counterparts.

Useful feature: bioRxiv integrates with many journal submission systems (eLife, PLOS, EMBO, BMJ, and others), allowing direct transfer from preprint to submission with one click.

Best for: Life scientists posting primary research, confirmatory replications, or negative results before or alongside journal submission.

3. medRxiv

URL: medrxiv.org
Disciplines: Clinical medicine, public health, epidemiology, health systems, global health, translational medicine
Operator: openRxiv / Yale University / BMJ
Cost: Free

medRxiv launched in 2019 and applies stricter screening than bioRxiv due to the potential public-health consequences of health misinformation. Screeners check that preprints do not contain unsupported claims that could harm public health, and the process may take two to three business days. The server explicitly warns readers on every preprint that the work has not been peer-reviewed.

Licensing: CC BY, CC BY-NC, or CC BY-ND options. Authors retain copyright.

Indexing: Google Scholar, CrossRef, Europe PMC, and a subset are indexed by PubMed Central. See the medRxiv FAQ for current indexing details.

Best for: Clinical trialists, epidemiologists, and public health researchers posting findings ahead of journal publication. Particularly valuable for rapid dissemination during disease outbreaks or when findings carry immediate policy relevance.

4. SSRN

URL: ssrn.com
Disciplines: Economics, law, finance, accounting, management, social sciences, humanities, and dozens of sub-networks
Operator: Elsevier
Cost: Free to post; free to download

SSRN (Social Science Research Network) is the dominant preprint platform for economics, law, and finance. It operates as a network of subject-specific repositories, each maintained by a scholarly society or editorial board. Researchers post working papers, draft articles, and book chapters. Unlike arXiv, SSRN papers are often posted at an earlier stage — sometimes before a submission is even planned — and revised repeatedly over years.

Screening: Minimal; papers are checked for format compliance and basic spam/integrity criteria. SSRN does not moderate scientific content.

Licensing: Authors retain full copyright. No open licence is required; by default papers are posted with standard copyright, though authors can choose CC licences.

Indexing: Google Scholar, CrossRef. SSRN papers receive a DOI upon posting.

Best for: Economists, lawyers, finance researchers, and social scientists who want working-paper distribution in the tradition of the field.

5. PsyArXiv

URL: osf.io/preprints/psyarxiv
Disciplines: All areas of psychology — clinical, cognitive, developmental, social, educational, neuroscience, and allied behavioural sciences
Operator: Society for the Improvement of Psychological Science (SIPS) / Center for Open Science
Cost: Free

PsyArXiv is the dedicated preprint server for psychology and the behavioural sciences, built on the OSF infrastructure. Launched in 2016, it has become a central venue for registered reports, replication studies, and open-science outputs in psychology. As of March 2026, moderation checks have been strengthened: moderators verify author credentials and scholarly expertise for review articles, case studies, and opinion pieces before public posting.

Licensing: Authors can choose any Creative Commons licence or retain copyright with no open licence. CC BY 4.0 is the recommended default for transparency and reuse.

Indexing: Google Scholar, CrossRef. DOIs are assigned at posting.

Best for: Psychologists who want to share preprints, especially alongside preregistrations or registered reports — both of which integrate seamlessly with the OSF platform. See also our guide on how to preregister a study on OSF and AsPredicted.

6. OSF Preprints

URL: osf.io/preprints
Disciplines: All disciplines (aggregator); hosts and federates more than 20 discipline-specific preprint servers
Operator: Center for Open Science
Cost: Free

OSF Preprints is both a general preprint server and an aggregation hub. It federates over 20 community servers — including PsyArXiv, SocArXiv, EarthArXiv, and others — into a single discovery interface, making it possible to search across more than two million preprints in one place. As a standalone submission platform, OSF Preprints accepts manuscripts from any discipline.

Beyond preprints, OSF is a full open-science workflow tool: you can link your preprint to a preregistration, an open dataset, analysis code, and supplementary materials in a single project space. This integration makes OSF especially powerful for transparent, reproducible research.

Best for: Interdisciplinary researchers, researchers in fields without a dedicated server, and anyone who wants to manage their entire open-science workflow (preregistration, data, code, and manuscript) in one place.

7. ChemRxiv

URL: chemrxiv.org
Disciplines: Chemistry and closely related areas — organic, inorganic, physical, analytical, computational chemistry, materials science, chemical biology
Operator: American Chemical Society (ACS), German Chemical Society (GDCh), Royal Society of Chemistry (RSC), and Chinese Chemical Society
Cost: Free

ChemRxiv launched in 2017 as the first dedicated preprint server for chemistry, operated by a consortium of leading chemical societies. Manuscripts undergo a basic screening for scope and research integrity. Papers are not peer-reviewed prior to posting. Authors can submit supporting data and supplementary files alongside the manuscript.

Licensing: CC BY, CC BY-NC-ND, or CC BY-NC-SA. Authors retain copyright.

Indexing: Google Scholar, CrossRef, SciFinder (Chemical Abstracts Service). DOIs are assigned at submission.

Best for: Chemists in academia and industry who want to establish priority and share results ahead of journal publication.

8. EarthArXiv

URL: eartharxiv.org
Disciplines: All subdomains of Earth science — geology, oceanography, atmospheric science, climate science, remote sensing, planetary science, geophysics, hydrology
Operator: Community-led; hosted by the California Digital Library (University of California)
Cost: Free

EarthArXiv is a community-driven preprint server launched in 2017, now hosted by the California Digital Library. It accepts manuscripts across all of Earth and planetary science. Submissions undergo scope moderation but no scientific peer review. All moderation decisions are transparent, and the server publishes its moderation policy openly.

Indexing: Google Scholar (crawled approximately weekly). EarthArXiv uses the OAI-PMH protocol so third-party indexers can harvest its metadata. DOIs are assigned at posting.

Best for: Geoscientists, climate researchers, and planetary scientists looking for rapid, community-validated dissemination without the fees and delays of traditional publishing.

9. Research Square

URL: researchsquare.com
Disciplines: Multidisciplinary — physical, biomedical, and social sciences
Operator: Research Square Company / Springer Nature partnership
Cost: Free to post

Research Square is a multidisciplinary preprint platform that operates across all scientific areas. It is tightly integrated with Springer Nature’s journal submission workflow through the “In Review” service: when you submit to a participating journal, your manuscript is automatically posted as a preprint on Research Square, making your work public while under review without any extra steps on your part.

Screening: Manuscripts are screened for scope and basic integrity, typically within two to four business days.

Indexing: CrossRef, Europe PMC, Google Scholar, Researcher, and Meta. A DOI is assigned at posting. PubMed Central indexes a limited subset. Source: Research Square indexing FAQ.

Best for: Researchers submitting to Springer Nature journals who want automatic preprint posting without managing a separate submission, and interdisciplinary researchers who do not have a single-discipline server.

General Repositories: For Data, Code, and Other Outputs

While preprint servers focus on manuscripts, general repositories accept datasets, code, figures, videos, presentations, and any other research output. Using a repository ensures your underlying data gets a DOI, is citable, and is preserved long-term — increasingly a requirement from funders such as UKRI, NIH, and Horizon Europe. For a full breakdown of funder requirements, see our guide to FAIR data principles for researchers.

10. Zenodo

URL: zenodo.org
Disciplines: All disciplines — truly general purpose
Operator: CERN / European Commission (OpenAIRE)
Cost: Free (up to 50 GB per record for the standard tier)

Zenodo is the go-to general repository for research outputs of all types. Funded by the European Commission and operated by CERN, it accepts datasets, software, presentations, reports, figures, videos, and manuscripts from any discipline. It has no requirements on format, access restrictions, or licence — though open licensing is strongly encouraged.

Every deposit receives a DataCite DOI and metadata is passed to DataCite and into scholarly aggregators, making outputs permanently citable. Zenodo integrates with GitHub: you can archive a specific release of a GitHub repository to Zenodo with two clicks, and it will assign a DOI to that version automatically — a common approach for citing research software.

Licensing: Zenodo’s own metadata is released under CC0 (public domain). Depositors can choose any licence for their content, though CC licences and open-source software licences are standard. The platform does not mandate any particular licence.

Indexing: DataCite, CrossRef, OpenAIRE, Google Scholar, and a broad range of academic aggregators. Zenodo records feed into the OpenAIRE system, which reports to EU funders.

Best for: Any researcher needing a free, long-term, DOI-backed home for data, code, or supplementary materials — especially those funded by Horizon Europe, which requires OpenAIRE-compatible deposit. Understanding what goes into a data availability statement alongside your Zenodo deposit is covered in our complete guide to writing data availability statements.

11. Figshare

URL: figshare.com
Disciplines: All disciplines
Operator: Digital Science
Cost: Free for up to 20 GB total storage (individual files up to 20 GB); Figshare Plus for larger datasets up to 5 TB

Figshare is a multidisciplinary repository that accepts any research output — datasets, figures, presentations, posters, code, videos, and preprints. Launched in 2011, it hosts more than eight million total research outputs. Many universities and publishers have institutional Figshare instances, making it a common mandated repository for institutional open-access policies.

Licensing: Code and software default to MIT, GPL-3.0, or Apache-2.0. All other file types default to CC BY 4.0. Institutions can mandate alternative licences for their institutional portals. DOIs are assigned at the file level and at the collection level, enabling version-controlled citations. Source: Figshare file limits documentation.

Indexing: CrossRef (via DOI), Google Scholar, Europe PMC for life-science content, DataCite.

Best for: Researchers whose institution has a Figshare portal, researchers who want browser-previewing of uploaded data, and anyone needing individual per-file DOIs for granular citation tracking.

12. Institutional Repositories

Most universities in the UK, US, Australia, Canada, and Ireland operate their own open-access repositories — for example, UCL Discovery, Edinburgh Research Archive, or the Australian National University’s research repository. Depositing here satisfies REF open-access requirements and UKRI mandate compliance for UK researchers.

When to use: When your funder or institution mandates deposit in an institutional repository, or when you want long-term preservation that is explicitly linked to your institutional affiliation. Most institutional repositories are harvested by OpenDOAR and BASE (Bielefeld Academic Search Engine), and their records feed into national research information systems such as PURE or Symplectic Elements.

How to find yours: The OpenDOAR directory lists over 6,000 open-access repositories worldwide, searchable by institution and country.

Comparison Table

How to Choose the Right Platform

Three questions will narrow your choice to one or two platforms in under a minute:

1. What type of output is it? A manuscript goes to a preprint server. A dataset, code archive, or supplementary file goes to a general repository. Both can go together — and should when your work has underlying data.
2. What is your field? Use the discipline-specific server where your community already looks: arXiv for STEM, bioRxiv/medRxiv for life and health sciences, SSRN for social sciences and law, PsyArXiv for psychology, ChemRxiv for chemistry, EarthArXiv for geosciences. Using the community server maximises early citations and feedback.
3. What does your funder or journal require? UKRI mandates immediate open access for articles and data. NIH’s Data Management and Sharing Policy requires a data management plan specifying a named repository. Check your journal’s preprint policy on SHERPA/RoMEO and your funder’s requirements before posting. The practical implications are explored in our article on research data sharing statistics, mandates, and compliance.

For tracking your scholarly identity across all these platforms, make sure your ORCID iD is connected to each record you deposit. Our guide to ORCID, DOIs, and persistent identifiers explains how to link preprints, datasets, and published papers to a single researcher profile automatically.

Once your data is deposited, you will need to reference it in your paper. A well-written data availability statement directs readers to the right repository and record. Templates for every scenario — open data, restricted data, data on request, and no data — are in our guide on how to write a data availability statement.

Frequently Asked Questions