The European Space Agency’s Euclid space telescope has delivered one of the most striking scientific results of the year. In its first major data release, called DR1-Foundation and published on 12 November, the mission revealed the largest catalogue of strong gravitational lenses ever assembled: 15,072 high-confidence candidates, more than all previously known catalogues combined. For a phenomenon that astronomers once counted in handfuls, that number marks a genuine turning point.
A strong gravitational lens forms when a massive foreground galaxy sits almost exactly between us and a more distant galaxy. The gravity of the nearer object bends space itself, and the light of the background galaxy is stretched, magnified and sometimes split into several images. In the most perfect alignments, the distant galaxy appears as a glowing circle around the foreground one, the famous Einstein ring predicted by general relativity more than a century ago.
Until recently, such systems were rare treasures. Each new find was a small discovery in its own right, often announced individually. Euclid has changed that. According to the team behind the catalogue, the new sample “transforms strong lensing from a field of rare discoveries to a robust statistical tool for astrophysics and cosmology.”
The scale of the release is hard to overstate. DR1-Foundation captures the light of more than a billion objects over an area of sky as large as 9,500 full Moons, and it is only the first step towards what ESA describes as the largest 3D map of the cosmos ever made. The lens search itself covered roughly 2,000 square degrees of that data.
Of the 15,072 high-confidence candidates, 6,073 are rated grade A, meaning the experts consider them “sure” lenses, while 8,999 are grade B, or “probable” lenses. The team also published a further 9,010 grade C candidates, systems that might be lenses but need more evidence. Around 2,580 objects that looked promising at first were ultimately judged not to be lenses at all.
How Machines, Volunteers and Experts Found Thousands of Cosmic Lenses
Finding these needles in a cosmic haystack required a new way of working. Among the roughly 900 million sources in DR1, the scientists selected about 68 million galaxies that met their criteria. No team of humans could ever look at 68 million images one by one, so the first pass was handed to artificial intelligence. Five different machine-learning models scored every galaxy for the likelihood that it was bending the light of something behind it.
The most promising candidates then went to people. Around 7,000 citizen scientists took part through the Space Warps project on the Zooniverse platform, where volunteers from all over the world inspected images and flagged the ones that showed arcs, rings or multiple images. Their classifications were combined with the machine-learning scores, and the strongest candidates were finally reviewed by 67 strong-lensing experts, who graded more than 26,000 systems in total.
This combination of algorithms, amateurs and professionals is increasingly common in astronomy, but rarely on this scale. The machines are fast and tireless, yet they can be fooled by spiral arms, galaxy mergers or image artefacts. Human eyes are slower but remarkably good at spotting the subtle, unusual shapes that make a lens special. Together, the approach maximised the number of lenses found while keeping individually valuable systems in the sample.
The sharpness of Euclid’s images was another key ingredient. In many cases, the team notes, the quality of the telescope’s visible-light camera is good enough to identify a system confidently as a strong lens from its shape alone. Definitive confirmation usually still requires spectroscopy, which measures the separate distances of the foreground and background galaxies, but Euclid’s pictures provide a very strong starting point.
Beyond the classic Einstein rings, the catalogue contains some genuine rarities. The researchers identified 84 candidate double-source-plane lenses, systems in which a single foreground galaxy bends the light of two different background galaxies lying at different distances. Such arrangements are extremely valuable because they allow scientists to test cosmological models with unusual precision. Only a few examples were known before.
The team also found more than 1,000 candidate disc lenses, in which the foreground lens is a flattened spiral-like galaxy rather than the usual giant elliptical. Again, this is a class with very few previously known members. On top of that, the catalogue includes several hundred candidate lensed quasars and active galactic nuclei, bright, energetic cores of distant galaxies whose flickering light can be used to measure the expansion rate of the universe.
Crucially, the scientists did not just publish a list. For all 15,072 grade A and B systems, they delivered automated lens models, along with estimated distances and stellar properties for the foreground galaxies doing the bending. That means other researchers can start using the data immediately, without first having to model each system themselves.
What do these models show? Most of the lensing galaxies are massive and “quiescent”, meaning they are no longer forming many new stars. Their typical stellar mass is around 250 billion times that of the Sun, under the standard assumptions used in the study. The lenses also tend to lie further away than those found in earlier ground-based surveys, with typical redshifts around 0.7, corresponding to light that left them roughly six billion years ago.
Why does all of this matter beyond the beauty of the images? Strong lenses act as natural scales for weighing galaxies. Because the amount of bending depends on the total mass of the foreground object, including invisible dark matter, lenses let astronomers map how dark matter is distributed in and around galaxies. With thousands of systems instead of dozens, researchers can now study how that distribution changes with galaxy type, mass and cosmic age.
Lenses are also powerful tools for cosmology. The magnified background galaxies offer a closer look at the early universe than would otherwise be possible, almost like a free zoom lens built into the cosmos. Lensed quasars, whose multiple images vary at slightly different times, can be used to measure the Hubble constant, the expansion rate of the universe, an area where different methods currently disagree.
The release is also a reminder of how much is still to come. DR1-Foundation is only the first major batch of data from a mission designed to survey a third of the sky over six years. Earlier forecasts suggested Euclid could ultimately find around 100,000 strong lenses, and the new catalogue suggests that goal is realistic. Each future release should add thousands more systems, along with ever more precise measurements of dark matter and dark energy.
For the thousands of volunteers who clicked through images on Space Warps, the result is a satisfying payoff. Their work is now part of the largest collection of cosmic lenses ever compiled, and the project shows that everyday people can still make a real contribution to frontier science.
Einstein himself doubted that gravitational lensing would ever be observed in practice. Today, a single telescope has turned his once-exotic prediction into a catalogue of more than 15,000 entries, and the hunt has only just begun.





