For years, mysterious little red dots kept appearing in James Webb images of the early universe — astronomers now think one is a “black hole star,” a growing black hole hidden inside a star-like gas cocoon the size of the solar system

A telescope worth ten billion dollars stared into the dawn of time and came back with freckles.

Within weeks of the James Webb Space Telescope starting science operations in 2022, astronomers began noticing them: tiny, fiercely red points sitting in almost every deep field. Someone called them little red dots, which was less a name than an admission of defeat, and it stuck.

Why the dots didn’t fit anything

Every available explanation failed in a different direction.

Read as young galaxies, the dots looked far too massive and too settled for a universe only a few hundred million years old. Cast instead as supermassive black holes gorging on gas, they were suspiciously quiet, missing the X-ray and radio racket that normally announces that kind of feeding. They were also abundant in the first billion years and largely gone by roughly two billion, which is not how populations of anything usually behave. As Scientific American put it, for a while the dots were framed as having broken cosmology.

One dot, in unusual detail

Rohan Naidu’s team named their survey Mirage or Miracle, on the suspicion that some of the impossibly bright early galaxies turning up in Webb images would prove to be the former. One source on their target list was redder and brighter than everything around it.

Naidu, then at MIT’s Kavli Institute for Astrophysics and Space Research and now at the University of Hawai’i, published the analysis with his co-authors in Nature in August. Their best fit is a black hole around 100,000 times the mass of the sun, wrapped in hydrogen so dense and so far-flung that the whole arrangement glows like one enormous star roughly the width of the solar system. In the description MIT published alongside the paper, the black hole does the work nuclear fusion usually does inside a star, and the gas enveloping it stands in for the star’s outer layers.

Nuclear fusion could never produce that kind of brightness. The object outshines the brightest known star by a factor of about 100 billion.

They named it MoM-BH*-1, the asterisk borrowed from black hole notation, the numeral a wager that others will follow. Its light left when the universe was roughly 660 million years old.

Red doesn’t always mean dust

Co-author Robert Simcoe, who directs MIT’s Kavli Institute, made the obvious point first: red usually means dust, in the same way that Canadian wildfire smoke turned Boston’s sky a strange orange. Dust filters out the shorter blue wavelengths and leaves the rest looking ruddy.

Except the spectrum had a cliff in it. Below a certain wavelength the light simply stopped, a feature known as a Balmer break that shows up in the atmospheres of middle-aged stars such as Vega. Naidu called it ‘the deepest break we have ever observed in any object’, too extreme for any ordinary star to produce. The spectrum also carried almost nothing heavier than helium, which is awkward for dust, since dust is built from heavier elements.

Other objects pointing the same way

Does one paper settle it? Not on its own, and the idea did not begin with this one. In September 2025, Anna de Graaff at the Max Planck Institute for Astronomy and colleagues reported an extreme dot they nicknamed The Cliff. They tried every existing model of a dusty galaxy or a dust-shrouded black hole against it, and failed each time. Their black hole star model was offered as a proof of concept, not a finished theory.

Two other results have pushed in the same direction. A Nature study in January argued that the broad emission lines in the best dot spectra are widened by electrons scattering light rather than by gas racing around under gravity, which would make the black holes involved something like a hundred times lighter than earlier estimates. In June, a team led by Vasily Kokorev at the University of Texas at Austin borrowed a foreground galaxy cluster as a magnifying glass. That delivered the deepest spectrum yet of a single dot, GLIMPSE-17775, with more than 40 spectral lines in it, most pointing to a black hole inside a hot, dense cocoon. As reported by the European Space Agency, Kokorev said part of the community is converging on a unified picture.

What the model still can’t explain

Nobody has photographed a black hole sitting inside a shell of gas, and nobody is claiming to have.

How the envelope forms, and how it survives while enclosing something that never stops eating, remain open questions. Rival explanations involving odd dust geometries or dense clusters of stars have not been ruled out. More Webb time is already booked.

Naidu has floated something bigger: that objects like this might set the pace at which galaxies can form stars in the first place, and everything that depends on stars eventually depends on them too.

One detail in the European Research Council’s summary of the work is easy to miss. MoM-BH*-1 sits next to a brighter galaxy, and the models suggest the two will merge in about 100 million years. Lay the two spectra on top of each other and the combination looks like an ordinary little red dot. Which means the tidiest example of the new class may look this clean only because it hasn’t been eaten yet.