Open access: what it is, who pays, and how to publish for free

What is open access?

Open access is the free, immediate, online availability of research, combined with the right to reuse it. The founding definition comes from the Budapest Open Access Initiative (2002): free availability on the public internet, "permitting any users to read, download, copy, distribute, print, search, or link to the full texts", to mine them and to use them "for any other lawful purpose, without financial, legal, or technical barriers other than those inseparable from gaining access to the internet itself."

The definition has two parts, which the philosopher Peter Suber later called gratis and libre open access:

• Gratis: nobody pays to read. Price barriers are removed.
• Libre: you may also reuse the work (translate it, teach with it, text-mine it, build on it) as long as you credit the authors. Permission barriers are removed, usually through a Creative Commons license.

Open access says nothing about quality: an open access article in a reputable journal goes through the same peer review as a subscription article. The idea also covers books, theses and data; this page focuses on journal articles, where the money questions matter most to researchers.

A short history: from Budapest to Bengaluru

Open access began as a researcher-led movement and became funder and government policy. The key milestones:

• 1991: the physicist Paul Ginsparg starts the preprint server that became arXiv, still the reference model for self-archiving.
• 14 February 2002: the Budapest Open Access Initiative (BOAI) defines open access and proposes two strategies: self-archiving in repositories (green OA) and open access journals (gold OA).
• 2003: the Bethesda Statement (20 June) and the Berlin Declaration (22 October) add a license allowing reuse with attribution and deposit in an online repository. Lund University launches the Directory of Open Access Journals (DOAJ) with about 300 journals.
• 2018: a group of research funders, cOAlition S, launches Plan S.
• 2021: UNESCO's member states adopt the Recommendation on Open Science.
• 2022: the BOAI20 recommendations call for a move away from article processing charges; Science Europe, cOAlition S, OPERAS and the French National Research Agency publish the Action Plan for Diamond Open Access (2 March).
• 2023–2026: Global Summits on Diamond Open Access meet in Toluca, Mexico (2023), Cape Town, South Africa (December 2024) and Bengaluru, India (February 2026). They produce the Toluca–Cape Town Declaration, which defines diamond OA as "community-owned, community-led, and non-commercial", and the Bengaluru Roadmap and Action Plan (May 2026). A fourth summit is planned in Bali, Indonesia, in December 2026.

Open licenses: what CC BY and the others allow

An open license turns "free to read" into open access: it tells everyone in advance what they may do with the work without asking. Almost all open access journals use Creative Commons licenses:

• CC BY (Attribution): anyone may copy, redistribute, adapt and reuse the work, including commercially, if they credit the authors. Plan S and Horizon Europe require it, and it is the most common license in DOAJ.
• CC BY-SA (ShareAlike): as CC BY, but adaptations must carry the same license.
• CC BY-NC (NonCommercial): reuse only for non-commercial purposes.
• CC BY-ND (NoDerivatives): sharing allowed, adaptation (including translation) is not.
• CC0: the authors waive their rights as far as the law allows; common for data.

Under open licenses authors usually keep their copyright. DOAJ's criteria state that "all rights reserved" is never appropriate for open access content.

The types of open access: diamond, gold, green, hybrid, bronze and black

The "colors" of open access describe where a paper is made open and who pays. They are conventions, not legal categories, and some sources treat diamond as a subtype of gold. For a deeper comparison of the three main routes, see our guide to diamond, gold and green open access.

Diamond (or platinum) open access

Diamond journals publish everything open access and charge neither readers nor authors. A university, research organization, learned society, library consortium or government covers the costs. Examples include SciPost Physics, the Journal of Statistical Software and Documenta Mathematica.

Gold open access

Gold journals are fully open access, but many fund themselves through an article processing charge (APC) paid by the author, the institution or the funder. PLOS ONE and Scientific Reports, for instance, list APCs of roughly US$2,500–2,900 in their DOAJ records.

Green open access (self-archiving)

The author deposits a version of the paper in a repository: a subject archive such as arXiv or PubMed Central, a national archive such as HAL in France, or an institutional repository. This is usually the preprint or the accepted manuscript (after peer review, before the publisher's layout). Publishers often impose an embargo first, commonly six, 12 or 18 months, although many funders now require zero embargo. Green OA costs the author nothing.

Hybrid open access

A hybrid journal is a subscription journal that opens individual articles if an APC is paid, while libraries keep paying subscriptions for the rest; critics call this "double dipping." Hybrid APCs are on average the most expensive: fully OA list prices average about two-thirds of hybrid ones, according to Delta Think. Nature is a hybrid journal.

Bronze open access

Bronze articles are free to read on the publisher's site but carry no open license, so reuse rights are unclear and the publisher can restore the paywall. The term comes from a 2018 study in PeerJ, which found bronze to be the most common form of free access at the time.

Black open access

"Black OA", a label popularized by Bo-Christer Björk in 2017, means copies distributed without permission by shadow libraries such as Sci-Hub and Library Genesis. They are widely used but infringe copyright in most countries; in 2017 a US court awarded Elsevier US$15 million in damages against them. We mention them for completeness, not as a recommendation: the legal routes below often lead to the same paper.

Comparison table

Who pays for open access?

Readers never pay; the cost moves to authors, institutions, funders or the community that runs the journal, depending on the business model.

Article processing charges (APCs)

An APC is charged once a paper is accepted, and prices vary enormously:

• Universities and funders reporting to OpenAPC (mostly European) paid on average about €2,400 per article in fully open access journals in 2024 and about €3,100 in hybrid journals (our calculation from the public OpenAPC dataset).
• Surveying more than 40 publishers, Delta Think found that in 2026 the highest list APC reached US$12,850 in hybrid and US$8,900 in fully OA journals, with list prices rising about 5–7% a year.
• A study in Quantitative Science Studies estimated that authors paid US$1.06 billion in APCs to the five largest commercial publishers between 2015 and 2018.

Transformative and "read and publish" agreements

Many libraries now negotiate deals in which subscription money also covers open access publishing by their authors; Germany's Projekt DEAL contracts with Wiley, Springer Nature and Elsevier are among the best known, and the ESAC registry collects such agreements worldwide. For an eligible author the APC is invisible. Critics, including the BOAI20 recommendations, argue that these deals entrench the largest publishers and leave authors at institutions without a deal, including most of the Global South, facing the full fee.

Waivers and discounts

Many publishers waive or reduce APCs for authors from lower-income countries. Elsevier, for example, grants a 100% waiver when all authors are at institutions in Research4Life "Group A" countries and 50% for "Group B", but its waiver policy "does not apply to hybrid journals." Waivers are discretionary and usually require every co-author to qualify.

How diamond journals are funded

The OA Diamond Journals Study commissioned by cOAlition S found that 60% of diamond journals depend on volunteers, 53% run on less than one full-time equivalent and 70% declared annual costs below $/€10,000. Support comes mainly from universities and research organizations, and also from:

• university presses and library publishing services running open-source software such as Open Journal Systems (OJS);
• platforms for non-commercial journals, such as Redalyc and AmeliCA in Latin America or OpenEdition in France;
• learned societies and library consortia that pool funds;
• Subscribe to Open, in which a journal opens each volume if enough libraries keep subscribing.

Diamond open access in numbers

Diamond journals are the majority of open access journals but publish a minority of open access articles. In the DOAJ snapshot this site uses (October 2026), 14,571 of the 23,657 listed journals (62%) charge no APC and no other fees. Of those, 4,279 are ranked by SCImago/Scopus, and 711 are in the top quartile (Q1) of at least one subject category.

They are also very international: the diamond journals in our dataset come from 132 countries and publish in 84 languages.

The most detailed study of the sector, the OA Diamond Journals Study (Bosman et al., 2021), found that:

• there are "almost certainly at least 17,000, but likely up to 29,000" diamond journals worldwide, only about a third of them registered in DOAJ;
• they published about 356,000 articles a year, around 8–9% of all scholarly articles, against about 453,000 (10–11%) in APC-funded OA journals;
• 45% are published in Europe, 25% in Latin America, 16% in Asia and 5% in the US and Canada;
• 60% are in the humanities and social sciences, 22% in the sciences and 17% in medicine, and most publish fewer than 25 articles a year;
• 38% are multilingual, against 14% of APC-based journals.

New infrastructure makes these journals easier to find. The Diamond Discovery Hub, launched in 2025 by the CRAFT-OA project and now part of the European Diamond Capacity Hub, passed 4,000 verified diamond journals in July 2026.

Open access policies and mandates in 2026

Most large public funders now require the papers they fund to be open from the day of publication; they differ on whether they pay publication fees and which license they demand. The situation in October 2026:

Plan S and cOAlition S

Plan S was launched on 4 September 2018 by cOAlition S, which now has 28 funders, among them Wellcome, the Gates Foundation, the World Health Organization, UK Research and Innovation, France's ANR and Jordan's Higher Council for Science and Technology. Since 2021, funded papers must be open on publication under CC BY, either in an open access journal or platform or through immediate repository deposit; funding for transformative arrangements ended on 31 December 2024, and the funders "do not support the 'hybrid' model". Their Rights Retention Strategy lets authors put a CC BY license on their accepted manuscript even when they publish in a subscription journal. The coalition's 2026–2030 strategy includes a review of the Plan S principles, with more weight on diamond open access, preprints and "publish, review, curate" models.

European Union and Horizon Europe

Horizon Europe grant holders must deposit each peer-reviewed publication in a trusted repository at publication, under CC BY or equivalent; hybrid and subscription journal fees are not reimbursed. They can publish free of charge on Open Research Europe, which from autumn 2026 is operated by CERN with nearly €17 million for 2026–2031 from the European Commission and 11 countries. In May 2023 the Council of the EU stated that immediate, unrestricted open access "should be the norm" for publicly funded research, with costs not covered by individual authors or readers.

UNESCO

The Recommendation on Open Science, adopted in Paris on 23 November 2021 by UNESCO's 193 member states, is the first global standard on the subject. It covers publications, data, software and infrastructure.

United States

The White House "Nelson memo" of 25 August 2022 told federal agencies to make funded papers freely available without embargo by 31 December 2025, and the NIH applied its zero-embargo policy from 1 July 2025. The direction has since changed. A 2026 congressional report states that the science policy office "is in the process of repealing" the memo (House Report 119-652); a proposed federal rule of 29 May 2026 would make APCs unallowable costs on federal grants unless required by statute or approved in advance; and NIH plans to cap publication costs. None of these had been finalized at the time of writing, so US-funded authors should check their agency's current rules. Repository deposit and diamond journals sidestep the cost question.

Elsewhere

• Canada: the revised Tri-Agency policy requires immediate open access under an open license (preferably CC BY) for grants awarded from 1 January 2027.
• India: One Nation One Subscription, running since January 2025, gives over 6,300 government institutions access to more than 13,000 journals and sets aside ₹150 crore a year for publishing in selected open access journals.
• Gates Foundation: since January 2025 its policy no longer pays APCs and expects preprints.

Benefits, criticisms and the citation advantage debate

Open access widens readership and reuse; its main problems lie in how it is paid for.

Benefits

• More readers. Practitioners, journalists, patients, teachers and researchers without large library budgets can read the work. In a randomized trial across 36 journals, open articles received significantly more downloads and reached a broader audience within a year.
• Reuse. CC BY allows translation, teaching material, systematic reviews and text-mining without permission requests.
• Public return on public money, the core argument of funder mandates.

Does open access increase citations?

Possibly, but the evidence is weaker than often claimed. A large 2018 analysis found that open access articles received 18% more citations than average. A 2021 systematic review of 134 studies, however, found that 47.8% confirmed a citation advantage, 27.6% found none and 23.9% found one only in subsets; only three studies had a low risk of bias, and they disagreed. The only randomized trial found open articles "cited no more frequently, nor earlier" within three years. Self-selection (authors open their best work) probably explains part of the advantage. Choose open access for reach and reuse, not guaranteed citations.

Criticisms

• APCs move the barrier from readers to authors. In a study of Elsevier's "mirror journals", most open access articles came from high-income countries and none of the open mirror articles came from low-income countries. BOAI20 says APC-based journals "exclude authors for economic reasons and without regard to the quality and importance of their work."
• Volume incentives. When income depends on accepted articles, publishers gain by accepting more. In 2023 Wiley's Hindawi journals retracted more than 8,000 articles, mostly from special issues exploited by paper mills.
• Predatory publishers. Some outlets take APCs without real peer review. A 2015 study estimated their output grew from 53,000 articles in 2010 to about 420,000 in 2014 in some 8,000 active journals, with three quarters of authors in Asia and Africa. See our guide to predatory journals.
• Fragile diamond finances. Reliance on volunteers and short-term grants can limit copy-editing, indexing and long-term preservation.

How to publish open access for free

The most direct way to publish open access without paying is a diamond journal that fits your paper and meets your institution's requirements. With the free tools on this site:

1. Build a shortlist. Browse the diamond journal finder by subject, country or language, or paste your abstract into the AI matcher, which runs on your device and compares it with each journal's scope.
2. Check the fit. Read the aims and scope and skim recent issues. A journal that published papers like yours in the last two years beats one with a vague scope.
3. Check the ranking. If your institution asks for a quartile, filter by SCImago quartile and read how quartiles work. Ready-made lists: Scopus-ranked journals with no APC and free Q1 journals.
4. Verify Scopus indexing by ISSN. Rankings are yearly and journals are sometimes discontinued; confirm the current status with the Scopus indexing checker and our step-by-step guide.
5. Rule out predatory signs. Check the DOAJ listing, a real and reachable editorial board, a described peer review process and no promise of acceptance within days. Use the predatory journals checklist.
6. Confirm "no fees". Look for submission, page or color charges in the author guidelines; a diamond journal states clearly that it charges nothing.
7. Submit, then keep a green copy. After acceptance, deposit your accepted manuscript in a repository and check which license your funder requires.

How to find open access versions of papers

A free, legal copy of a paywalled article often exists in a repository or on an author's page. Try:

• Unpaywall: a free browser extension that flags legal open copies of the article you are viewing.
• Google Scholar: click "All versions" under a result to find repository copies.
• CORE and BASE: search engines aggregating full texts from thousands of repositories and journals.
• Subject archives: PubMed Central and Europe PMC for life sciences, arXiv for physics, mathematics and computer science, bioRxiv, medRxiv and SSRN for other preprints.
• National and institutional repositories, such as HAL in France.
• Ask the author. Researchers are usually happy to email a copy of their paper.

Frequently asked questions

Is open access peer reviewed?

Yes, in legitimate journals. Open access is about distribution, not evaluation. DOAJ, for example, only lists journals in which every article is reviewed by at least two independent reviewers (or two editors in the arts and humanities). Preprints are the exception: they are shared before review and labeled as such.

Is open access free to publish?

Not always. Diamond journals and green self-archiving are free for authors; gold and hybrid journals usually charge an APC, which your library or funder may cover. In DOAJ, 62% of journals charge nothing; you can filter them in the finder.

What is the difference between open access and free access?

Free access (sometimes called public access or bronze OA) means you can read the article at no cost. Open access adds an open license that permits reuse, and it is permanent: a publisher can withdraw free access at any time but cannot revoke a CC BY license already granted.

Are open access journals indexed in Scopus?

Many are. Scopus selects journals on quality criteria, not business model; 4,279 diamond journals in our dataset have a SCImago/Scopus ranking. A DOAJ listing does not imply Scopus indexing, so always check the ISSN.

What does CC BY mean?

CC BY is the Creative Commons Attribution license. Anyone may share, adapt and reuse the work for any purpose, including commercially, if they credit the authors, link to the license and indicate changes. Most funder open access policies require it.

Are open access journals predatory?

No. Predatory journals are a minority exploiting the APC model; most open access journals, including the reputable diamond journals in DOAJ, run normal peer review. Check the warning signs in our predatory journals guide.

Can I make a paper open if I published it in a subscription journal?

Usually, yes, through green OA. Most publishers allow deposit of the accepted manuscript, sometimes after an embargo; check the journal in Open Policy Finder (formerly Sherpa Romeo). Some countries, such as France and Germany, also give authors a statutory right to do so.

Sources

• Budapest Open Access Initiative (2002); BOAI20 recommendations (2022); Bethesda Statement and Berlin Declaration (2003)
• Bosman et al. (2021), The OA Diamond Journals Study, Part 1: Findings
• Action Plan for Diamond Open Access (2022); Toluca–Cape Town Declaration (SPARC, 2025); Bengaluru Roadmap (Science Europe, 2026); Diamond Discovery Hub milestone (EDCH, 2026)
• DOAJ, criteria for inclusion and DOAJ at 20
• Piwowar et al. (2018), The state of OA, PeerJ; Langham-Putrow et al. (2021), Is the open access citation advantage real?, PLOS ONE; Davis (2011), randomized controlled trial, FASEB Journal
• Smith et al. (2021), APCs and author geographic diversity; Butler et al. (2023), The oligopoly's shift to open access, both Quantitative Science Studies
• Shen & Björk (2015), 'Predatory' open access, BMC Medicine; Björk (2017), Gold, green, and black open access, Learned Publishing
• Delta Think (2026); OpenAPC data; Elsevier waiver policy
• Nature news: retractions in 2023; Elsevier v. Sci-Hub (2017)
• cOAlition S: Plan S principles, Rights Retention Strategy, Strategy 2026–2030
• ERC open science rules; Open Research Europe (2026); Council conclusions (2023); UNESCO Recommendation (2021)
• US: OSTP memo (2022); House Report 119-652; OMB proposed rule (2026); AIP FYI (2026)
• Canada Tri-Agency policy (2026); India ONOS (2025); Gates Foundation policy
• Suber (2008), Gratis and libre open access; Emery (2018), embargo periods, Insights