A preprint platform is paying researchers for reviewing preprints uploaded by other scientists, reigniting debates about whether peer reviewers should be paid.

XSci, which launched in September, will reward a reviewer with US$200 if they send their feedback within five days of accepting the task and $100 if they do so between five and ten days. The fee is covered by other researchers, who pay XSci to have their preprints reviewed by scholars. The preprint website, which does not publish a journal, charges each author $399 for a round of review or $249 if they make their review report public.

The system enables authors to receive comments quickly while compensating reviewers for their time and effort, says Shumiao Ouyang, the founder of XSci, the start-up firm incorporated in Delaware that runs the platform. The company has received around $500,000 from private investors and is trying to raise more funds, Ouyang says.

A publishing platform that places code front and centre

“What authors get is an answer, usually within about five to ten days, to the question they care about most: does this paper have a real chance at the journals I am aiming at? Today, that answer can take months, and in some fields, years,” says Ouyang, who is an economist at the University of Oxford, UK. XSci also charges authors up to $19 to have their preprints reviewed by a range of artificial-intelligence programs, although access to some AI tools is free.

Several researchers interviewed by Nature say that XSci’s business model is interesting, but that its future will depend mostly on whether it can attract authors to use its human-reviewing service.

“Will authors be willing to pay such a fee? Where does the money come from?” asks developmental biologist and toxicologist Daniel Gorelick at Baylor College of Medicine in Houston, Texas, who has served as editor-in-chief of Biology Open since July 2023.

But Gorelick also says that cash payments are “a powerful incentive”. Biology Open, an open-access journal published by The Company of Biologists, has been paying reviewers £220 (or $280) per manuscript since 2024 through a programme called Fast and Fair. The payment is issued to the reviewer if they send in their work on time and if its quality meets the editors’ expectations.

Ouyang expects XSci’s core authors to be doctoral and postdoctoral students, as well as early-career researchers “who often cannot wait for a year or more for a journal’s decision”.

Peer reviewers are conventionally expected to review manuscripts as a service to the scientific community, rather than for a fee. By paying them at the preprint stage, Ouyang says, XSci intends to tackle challenges that dog the academic publishing industry today: how to convince scholars to review the work of others and to do it quickly.

Several studies^1–3 have shown that journal editors are finding it increasingly difficult to source reviewers for various reasons, including a rising number of submissions and academics’ growing workloads. These and other issues often lead to a long review time, which can range from a few weeks to several months.

XSci states on its website that academics can register their interest to be paid reviewers. But in reality, says Ouyang, a scholar must be invited by the website to review a manuscript. The website nevertheless has a ‘register interest’ field for practical reasons because, he explains, “it tells us how many willing researchers there are in each field.”

Ouyang says that XSci has developed an AI program to help find suitable reviewers quickly. Once an author uploads their manuscript to XSci and chooses the human-reviewing option, the AI tool will compile a ranking of potential reviewers by screening candidates from three sources: the scientists who wrote the papers that the manuscript cites, those scholars’ regular collaborators and researchers who have published papers on similar topics in the author’s target journals. The AI program will rank, then e-mail, potential reviewers one by one. Each potential reviewer has 12 hours to accept the job, says Ouyang.

If the reviewer thinks that the manuscript could be published in the author’s target journals after reasonable changes, the preprint will be given a badge on the website, demonstrating that it has been reviewed and is endorsed by the reviewer.

In time, XSci hopes to form partnerships with journal publishers that would allow the journals’ editors to use XSci’s review reports as part of their assessments. This would happen only if the authors and reviewers both agree to the reports being shared, Ouyang says. The idea is that editors can decide what to do with XSci’s reviews: whether to use them to replace one round of their own reviews or as external references, he adds. In follow-up questions from Nature, however, Ouyang stressed that the badge would merely “become useful information for an editor”.

“Whether a journal gives them any weight is the entirely the journal’s decision,” he added.

Pros and cons

In 2025, Fast and Fair helped to slash the time that Biology Open editors spent making the first decision about a manuscript from an average of 37.7 working days under the free-reviewing route to an average of 5.5 working days, according to a preprint co-authored by Gorelick⁴. The programme also led to review invitations having a higher acceptance rate and a higher completion rate of accepted tasks, the study found.

But a separate study found that paying reviewers did not achieve substantial improvements. Editors at Critical Care Medicine, a journal published by Lippincott Williams & Wilkins, carried out a six-month experiment in which they sent reviewers invitations with or without monetary incentives in alternating two-week blocks⁵. They found that 50% of review invitations with a $250 cash reward resulted in a peer-review report in 14 days, compared with 42% of invitations without the fee. Paid reviewers also delivered their work just one day quicker on average than did non-paid reviewers.

“It didn’t move the needle too much,” says David Maslove, one of the editors who co-authored in the study and a clinician-scientist at Queen’s University in Kingston, Canada.

‘Getting paid to review is justice’: journal pays peer reviewers in cryptocurrency

A paying system comes with challenges, researchers warn. There need to be stringent quality controls, as well as policies in place to ensure that reviewers don’t game the system and publishers don’t take advantage of reviewers, Gorelick says. “You can’t just pay reviewers and walk away,” he notes.

“XSci currently does not check the quality of the reviews done by the reviewers before they are released to the authors,” says Jeffery Poon, publishing and editorial lead of XSci. If the authors are not satisfied with what they receive, they can initiate an appeal through XSci, Poon notes. XSci plans to provide tools to reviewers to help them check their own work, but the team is still working out how that could be implemented in reality, he adds.