Science publishing is expanding across several measurable dimensions at once: research output, open-access availability, digital readership, metadata infrastructure, and integrity checks. The figures below describe different organizations and measurement periods, so they should be read as clearly labeled snapshots rather than one unified industry total.
Contents
- Open-access publishing
- Research output and readership
- Submissions, decisions, and integrity
- Disciplinary differences in open access
- The DOI and metadata infrastructure
- Authors, editors, and reviewers
Open-access publishing
Springer Nature reported that 50% of its primary research was published open access in 2024. Its 2024 Open Access Report also recorded 240,000 open-access articles, a 31% increase from the previous year. These figures describe Springer Nature’s publishing activity, not the entire scholarly publishing market.
The same report indicates that transformative agreements powered 10 times more gold open-access articles in 2024 than in 2023 at Springer Nature. Gold open access means the article is made openly available by the journal, commonly through an article processing charge or an agreement that covers publishing costs. At Springer Nature, open-access journal article downloads rose 30% year over year in 2024, while open-access book downloads rose 31%.
The cost side is also visible in Springer Nature’s publishing data. The company said it waived more than EUR 26 million in article processing charges in 2023. More than EUR 8.5 million of that support went to authors in World Bank low-income and lower-middle-income countries or regions. These are reported waiver amounts, not a measure of all publishing costs faced by researchers.
The STM OA Uptake by Discipline figures provide a broader disciplinary comparison for 2024. Gold open access was available for 81% of STM publications and 80% of SSH publications. When gold open access was offered, STM authors chose it 50% of the time and SSH authors chose it 52% of the time.
| Measure | STM | SSH |
|---|---|---|
| Share of global articles, reviews, and conference papers in 2024 | 90% | 10% |
| Gold OA available in 2024 | 81% | 80% |
| Authors choosing gold OA when offered in 2024 | 50% | 52% |
| Gold OA availability in 2014 | 58% | 29% |
Source: STM OA Uptake by Discipline. STM refers to science, technology, and medicine; SSH refers to social sciences and humanities. The source reports that the share of gold open access increased 26% in STM and 32% in SSH over the ten years to 2024. It also reports that the share of subscription-only articles fell 15% in STM and 27% in SSH over that period.
Research output and readership
Springer Nature’s 2024 Annual Report recorded 482,000 primary research articles and 14,000 books published during the year. It reported EUR 1,847 million in revenue and EUR 512.4 million in adjusted operating profit. The annual report also recorded 3.7 billion content downloads across all platforms, more than 10 million downloads per day, and 18% growth in downloads in 2024. Its reported average was 5.2 citations per article, up 4%.
The annual-report figures show how different publishing indicators answer different questions. Article counts measure output, downloads measure access or use, citations measure subsequent scholarly attention, and revenue measures the publisher’s financial scale. None of these metrics alone describes research quality or public impact.
Springer Nature’s value-of-publishing information says it publishes over 420,000 articles a year across more than 3,000 journals and over 14,000 research books a year. That annual figure is presented as a publishing scale indicator and is not identical in scope to the 482,000 primary research articles reported for 2024.
For the Springer Nature Link platform specifically, the 2024 technology and AI reporting recorded more than 900,000 registered users, 20 million active users each month, and 185 million monthly downloads. Across all platforms, Springer Nature reported more than 2 million registered users. These user counts use different definitions, so registered users should not be added to monthly active users.
Elsevier’s publishing cycle deck reports a different large-scale publishing snapshot: 30 million researchers served, more than 480 million downloads per year, 2,000 journals, 450,000 new articles each year, and 185 years of back issues scanned, processed, and data-tagged. It also reports 7,000 editors, 70,000 editorial board members, 500,000 reviewers, and 7 million author or publisher communications per year.
Submissions, decisions, and integrity
Submission volume provides context for published output. Springer Nature’s Research Integrity page says it received over 2.3 million submissions in 2024 and published only 482,000 or more of them. The same source records 2,923 retractions in 2024. A retraction is a post-publication integrity action; it should not be treated as a rejection rate.
Of Springer Nature’s 2024 retractions, 61.5%, or 1,797 articles, were for papers published before January 2023. The remaining 38.5%, or 1,126 articles, were for articles published after January 2023. Among the retractions for content published after January 2023, 41% were for articles in open-access journals. These figures describe retractions carried out in 2024, not the rate at which all 2024 publications were later retracted.
The same Research Integrity source reports a later snapshot for 2025: over 3.1 million submissions, 539,000 articles published, and 1,462 retractions. In that year, 57%, or 833 articles, of the retractions concerned papers published before January 2024, while 43%, or 628 articles, concerned articles published after January 2024. The 2025 figures are a separate measurement period and should not be combined with the 2024 totals.
Elsevier’s deck gives a range rather than one overall acceptance figure: it says Elsevier rejects 40% to 90% of articles. Its example for BBRC records 8,000 manuscripts submitted per year, a 25% acceptance rate, 29 editors, and a 10-day submission-to-decision time. The BBRC numbers are an example in the deck, not a reported acceptance rate for every Elsevier journal.
Springer Nature also reported more than EUR 177 million in technology-related spending, including research integrity, in 2024. Its annual report said there were more than 90 live AI-enabled processes and solutions by the end of that year. The report describes Snapp, its publishing platform, as having supported more than 12 million authors, handled more than 2 million submissions, published more than 390,000 articles, and worked with 78,000 editors since launch.
Disciplinary differences in open access
The disciplinary data show that availability and author choice are separate measures. In 2014, gold open access was available for 58% of STM publications and 29% of SSH publications. By 2024, availability had reached 81% in STM and 80% in SSH. The source reports increases of 26% in STM and 32% in SSH over the ten-year period.
Author behavior did not equal availability in 2024. STM authors chose gold open access half the time when it was offered, while SSH authors chose it 52% of the time. This distinction matters because a journal can make an open-access route available without every eligible author selecting it.
The same source places the disciplines in context: STM accounted for 90% of global articles, reviews, and conference papers in 2024, while SSH accounted for 10%. The comparison therefore combines a large difference in publication volume with similar 2024 rates for gold open-access availability and author choice.
The DOI and metadata infrastructure
Crossref’s statistics page reported 183,269,232 total records and 34,856 prefixes on May 11, 2026. It listed 163,709 journals and 139,625 conference proceedings. Its record types included 123,504,176 journal DOIs, 9,880,213 conference DOIs, 9,317,228 components, 505,789 standards titles, 70,435 databases, and 199,944 grant DOIs.
Crossref also reported 2,005,894,912 cited-by links on that measurement date. Its title-level DOI counts included 81,448 journal titles, 2,427,687 book titles, 82,571 conference proceeding titles, 1,061,310 dissertations, 951,927 reports, 413,702 standards titles, and 58,199 database titles. It recorded 2,947,660 unique DOIs with FundRef data.
| Crossref activity | 2024 | 2025 |
|---|---|---|
| Records added | 10,886,577 | 12,137,579 |
| Current deposits | 9,424,737 | 10,302,272 |
| Backfile deposits | 1,461,840 | 1,835,307 |
| Total records updated | 53,544,428 | 78,775,016 |
Source: Crossref stats page. Crossref reported 1,868,141,440 metadata queries with 920,105,664 matches in 2024, and 2,047,785,984 metadata queries with 1,257,433,216 matches in 2025. DOI queries totaled 3,229,297,408 in 2024, with 3,038,507,264 matches; in 2025, DOI queries totaled 3,302,196,736, with 3,001,981,952 matches.
Crossref’s ORCID-related figures included 1,690,589 works added by its auto-update system, 40,231,320 ORCID/article pairs, 5,809,732 authors granting permission, 12,355 authors denying permission, and 3,592,881 authors not responding. These are infrastructure and participation measures, not counts of unique researchers across the entire scholarly ecosystem.
Authors, editors, and reviewers
The scale of human participation is large alongside the digital infrastructure. Springer Nature’s value-of-publishing page reports more than 174,000 academic editors and more than 1.2 million peer reviewers in its publishing workflow. It says Springer Nature evaluates more than 1.8 million submissions a year. It also reports more than 420,000 articles a year across more than 3,000 journals and over 14,000 research books a year.
Springer Nature’s 2024 Early Career Researcher programme mentored 25 researchers. The annual report says those mentors came from minority-serving institutions and underserved communities across 11 countries. This is a programme-specific participation figure, not a measure of all early-career researchers supported by the publisher.
Together, these statistics describe science publishing as a connected system: journals and books produce research, open-access routes change availability, platforms measure readership, editors and reviewers mediate decisions, integrity teams handle post-publication corrections, and DOI registries make scholarly objects discoverable and linkable. Each number remains tied to its named source, scope, and measurement period.