By the time a star in Omega Centauri finishes a single loop around its companion, a person born at the start of that orbit would be pushing a hundred. The star takes 94 years to come around once, and the thing it circles gives off no light at all. After weighing the star’s slow arc across two decades of images, astronomers have concluded that the dark companion is a black hole of about four and a half times the Sun’s mass. It is the first stellar-mass black hole ever pinned down in Omega Centauri, and the pair holds the longest orbital period of any black hole binary yet found.

The discovery, reported in a new paper in The Astrophysical Journal Letters, came from combining more than twenty years of Hubble Space Telescope images with recent observations from the James Webb Space Telescope. The black hole has been named oMEGACat BH-2. Its real significance is as the first sighting of a population that models insist should be there but that has stubbornly refused to show itself.

A black hole the cluster was supposed to be hiding

Omega Centauri is the largest and most massive globular cluster orbiting the Milky Way, a tight swarm of roughly ten million stars about 18,000 light-years away. Clusters like it are old, and the most massive stars in them burned out long ago. Many should have collapsed into stellar-mass black holes, the kind left behind when a big star dies, weighing a few to a few tens of times the Sun.

Models of Omega Centauri suggest it should hold something like ten thousand of these black holes. For decades almost none turned up. Searches looked for the X-rays and radio waves a black hole gives off when it is pulling gas off a neighbor, and they scanned for the telltale wobble of stars being tugged by unseen weight. The black holes stayed dark and silent. Astronomers had already found evidence of a single much larger intermediate-mass black hole near the cluster’s center, but the far more numerous small ones stayed missing.

How you photograph something invisible

The new detection leaned on a different technique, called astrometry, which means measuring the tiny shifts in a star’s position over time rather than its light or its radio glow. A star bound to a heavy dark object does not sit still. It swings around the pair’s shared center of mass, and if you can track that swing precisely enough, the weight of the invisible partner falls out of the math.

Precision was the whole game. The team followed one star through 351 Hubble exposures reaching back to 2002 and added fresh Webb frames from 2024 and 2025, a record spanning about 23 years. The motions involved are smaller than a single pixel on the telescopes’ detectors. “The precision of these measurements is incredible, down to a fraction of a pixel,” said lead author Matthew Whitaker of the University of Utah in NASA’s announcement of the find, noting the black hole could not have been found without both telescopes.

The luminous star is an ordinary one, near the point where cluster stars begin to age off the main sequence, and weighs about 0.78 times the Sun. Once its mass was fixed, the orbit gave up the dark companion at roughly 4.46 solar masses. That is too heavy to be a neutron star, the dense stellar corpse that is the only real alternative, which tops out near two solar masses. An earlier group that spotted this star’s motion had guessed its hidden partner was a neutron star, and the longer baseline plus the Webb data ruled that out.

A 94-year orbit, and why it matters

The orbit itself is the record-setter. The visible star’s path around the pair’s center of mass is about 31 times the size of Earth’s orbit around the Sun, and it takes 94 years to complete, far longer than any black hole binary previously clocked. For comparison, the two stellar-mass black holes confirmed earlier in another cluster sit in orbits measured in days to months.

That wide, slow, lopsided orbit is a clue to how the pair came to be. The researchers think the two objects formed separately and only later found each other, shoved into partnership by the crowded traffic of the cluster core, a process called dynamical formation. Such a loosely bound pair is fragile. The team calculates it will likely be pulled apart by passing stars within less than a billion years, a blink against the cluster’s age of roughly 12 billion years.

The black hole’s modest weight is its own puzzle. Omega Centauri is metal-poor, meaning its stars formed with very little of the heavier elements, and theory tends to predict such environments should breed heavier black holes. A 4.46-solar-mass black hole here suggests at least some low-mass black holes can form even in these bare conditions. Coauthor Anil Seth called the result “surprising and exciting,” adding that “we now know that a metal-poor star is able to form a black hole like this, and we need to figure out how that happens.” The stakes reach past one cluster: dense clusters like this are thought to be where black holes pair up and eventually merge, sending out the gravitational waves now being caught on Earth.

An orbit caught before it finishes

The strongest caveat sits inside the headline number. Twenty-three years of watching covers well under half of a 94-year orbit, so the period is really a best estimate drawn from a partial arc. The paper’s own range runs from about 52 years to 157 years at one standard deviation, and the detection worked at all only because the observations happened to catch the star during its fast, close swing past the black hole around 2012. Away from that moment the star barely moves.

The mass carries similar hedges. The 4.46-solar-mass figure depends on the assumed 0.78-solar-mass weight of the visible star, which shifts if that star turns out to be enriched in helium, as a large fraction of cluster stars are. The team judges enrichment unlikely for this particular star, but it is a live source of error. A thin sliver of the data could in principle still permit an unusually heavy neutron star rather than a black hole, though the authors expect a couple more years of Webb measurements to close even that gap.

The ten-thousand figure deserves the same care. It comes from models of the cluster’s dynamics, not a headcount, and the authors themselves caution it is likely an overestimate, partly because earlier models did not fully account for the big central black hole. What the new work delivers is not the population but the first confirmed member of it, measured one star at a time.

What comes next

The team frames oMEGACat BH-2 as a starting point rather than a conclusion. Their models say more such binaries should be lurking in the same cluster, and further Webb observations should sharpen this orbit and settle the last neutron-star question. Further out, they are counting on NASA’s Nancy Grace Roman Space Telescope, which will survey crowded star fields on a regular cadence with Hubble-sharp vision.

Finding this black hole took two telescopes trained on the same speck of light for 23 years, and a single lucky stretch when the star swung close and fast. The next ones will ask for the same patience, and the instruments to spend it are only now coming online.