The best science in the world is being written right now. Most of it, you’ll never hear about.
Every year, hundreds of thousands of researchers wake up early, stay up late, and skip meals to pour their focus into understanding the world at its most fundamental level. They run experiments that fail dozens of times before they work. They spend years chasing questions most people don’t even know to ask.
And for every researcher struggling to be heard, there’s another struggling to find what they need; a scientist trying to build on work they can’t surface, a clinician unable to trust in findings they can’t evaluate, or an industry making high-stakes decisions on a literature nobody has properly mapped.
What connects all of them is a single broken system. A research paper sits in a queue, gets read by two or three people, and either disappears into a slush pile or emerges months or even years later, assessed not purely on what it says, but also on who said it and where they work.
A problem worth getting angry about
Journal rank is the way the scientific community assigns worth to research today, but journal acceptance rates are far from being scientific. It’s about where you’re from, as researchers achieve a 28% higher acceptance rate and higher overall reviewer scores when prestigious names and institutions are visible. It’s about who you are, as the median time under review is 7.4%–14.6% longer for female-authored manuscripts than for male-authored ones. And it’s about the average prestige of the journal, not specifically about the science or research of a single paper.
Meanwhile, the discovery that could accelerate someone else’s research, the finding that could connect two dots in a lab on the other side of the world, the insight that might shorten the path to a drug, a diagnosis or a breakthrough is lost. And the scientist behind it doesn’t get the recognition, promotion, or reward they deserve.
This is the system we’ve built to celebrate and share humanity’s best thinking.
What if the science could speak for itself?
Surfacing the best science, losing the prestige bias, and significantly accelerating the existing process would benefit the whole scientific community, and this is why we built QED Score. Our AI validity engine strips away the author’s name, their institution, the journal being targeted, and even the phrasing of the work, and asks, “Is this science valid? Original? Exceptional?” And it does it in minutes instead of months.
After validating our score across three rigorous studies, we felt confident to put it to its most challenging test yet; The 1%. 57,455 preprints assessed. A full year of life science output spanning May 2025- May 2026, ranked by the quality of the science alone. The 1% of that corpus is work more selectively identified than the 6-8% of submissions Nature or Science publishes, and it surfaces researchers who may never have had a platform like this before.
And the data behind it only deepens the story:
Nearly six in ten of the most exceptional papers involve international collaborators. Science, it turns out, travels across borders in ways that can’t be replicated by simply hiring more people at home.
The funders backing the best work aren’t always the ones you’d expect: a small New York philanthropic fund and a pediatric cancer charity outperform bigger funders on quality rate by more than a factor of ten.
And exceptional science doesn’t belong to any single geography. A compact cluster of institutes in Vienna, large domestic teams in China, and the vast distributed enterprise of American research all arrived at the top tier by entirely different routes.
Why preprints?
None of this would have been visible without looking at the right place, at the right time.
In science, standing on the shoulders of giants is more than just a cliche. Science is cumulative by nature, and each finding only matters if it reaches the people who can use it. That means when important work sits unpublished, or buried in a preprint nobody found, or locked behind a journal paywall while the review process crawls along at its own pace, the cost isn’t just to that researcher’s career. It’s to everyone whose work depends on theirs.
That’s why we targeted preprints for The 1%. 1.5 million preprints are submitted in biomedicine and life sciences every year. The signal-to-noise ratio for that volume of papers is brutal, and most researchers simply don’t have time to find the gems. But the potential benefit of gleaning the treasure from the corpus is immense, allowing others to build on findings months or even years earlier than they could otherwise.
QED Score makes that possible for the first time. It is a true quality signal, available the moment a manuscript exists. Not a replacement for peer review, but a way to surface the work that deserves attention before the slow machinery of publishing gets around to it. In many ways it can revolutionize how we sift through the scientific corpus in a similar way to how Google’s page ranking changed everything about navigating the Web. The researcher in Lagos or Lahore with a genuine breakthrough doesn’t have to wait for the right editors to notice. The work can be seen now via QED Science’s paper review and Insights Feed.
Behind every paper in The 1% is a researcher who deserves the recognition, but may never have expected it due to a broken system. Science is a deeply human act, powered by curiosity, stubbornness, creativity, and a willingness to be wrong, and the people doing it deserve better than a system that could filter them by zipcode before it ever reads their work.
The ambition of The 1% isn’t just to build a better ranking system. It’s to embed recognition into the scientific process. QED Score can provide that recognition, saying, “this work is exceptional on its own merit, period.” If you’re in The 1%, this is CV capital, fodder for the tenure committee, the headline of your next grant application. It says, this science was assessed by its originality and validity, and it was found to be among the best in the world.
The commercial benefit of QED Score
While The 1% is a badge of honor, the 99% is data.
The score isn’t just a celebration of the best. It’s a quality layer on top of an entire corpus, and every data point in that layer has commercial value to anyone making high-stakes or consequential decisions based on the scientific literature. Knowing that a study your drug pipeline depends on sits in a field where QED Score shows the weakest correlation with expert judgement; that’s a flag worth having.
Now think wider. The pharmaceutical company deciding which biological target to prioritize. The biotech reading the landscape for where the real breakthroughs are happening. The R&D team that needs to know whether the study their pipeline depends on is standing on solid ground or received wisdom. These industries have always run on science. Now, for the first time, they can see it ranked honestly, quickly find what matters, trust it, and move on it.
The science that changes everything
A preprint posted today carries no signal of its quality for months, sometimes years, while it waits in a review queue. The signal it eventually gets, where it was published, says more about the venue than the science. QED Score gives that signal on day one: a claim-level assessment of originality and validity, generated from the manuscript alone, before institutional reputation or journal placement has had any chance to shape how the work is read.
By organizing the scientific literature into a trusted, ranked, and navigable body of knowledge, we can compress the distance between discovery and application. The science exists, and the tools to surface it, evaluate it, and connect it to the people who need it exist. What remains is to build the layer that makes it all legible, and to make that layer available to every researcher, clinician, and decision-maker who depends on science to move forward.
About QED Science
QED Science is an AI company on a mission to surface the best science, and celebrate the scientists behind the manuscripts. Built on the belief that scientific quality should speak for itself, QED Science has developed QED Score, a validated quality metric that evaluates life science manuscripts on originality and validity after full anonymization, removing the influence of author reputation, institutional prestige, and publication venue. In use by more than 10,000 laboratories across 1,500 institutions in over 70 countries, QED Score delivers assessments in minutes that typically take months. Beyond discovery, QED’s validity engine empowers researchers to understand which claims hold up across papers, grants, and their day-to-day scientific work, without training on proprietary data. The 1%, QED Science’s selection of the most exceptional life science preprints, is the largest blind quality assessment of preprint science conducted to date.
