JWST’s little red dots are black holes feeding at 10 times the allowed rate

Since 2023, astronomers have been puzzled by a class of objects James Webb Space Telescope keeps finding scattered across the early universe: small, strikingly red, too bright for their apparent size. They called them “little red dots,” and the name stuck partly because nobody could explain what they were.

A new theoretical model proposes an answer, and it is more dramatic than anything previously suggested. According to a preprint by astrophysicists Yangyao Chen and Houjun Mo at Nanjing University and the University of Massachusetts, the little red dots are young supermassive black holes — masses between 100,000 and a million suns — undergoing episodes of accretion so violent they feed at rates up to ten times the theoretical maximum. The Eddington limit, the point at which a black hole’s own outward radiation pressure should prevent further mass from falling in, appears to be more of a guideline than a hard cap.

A mystery three years in the making

When JWST began returning its first deep-field images in 2022 and 2023, the little red dots weren’t in any catalog. Compact, faint, and redder than expected for their redshifts, they appeared in astonishing numbers for objects at the cosmic dawn — the first billion years after the Big Bang.

Early proposals were wide-ranging: a new class of exotic star, dense pockets of dust, or even a fundamental gap in the standard cosmological model. One competing explanation published earlier in 2026 proposed that apparent masses are being overestimated by a factor of 100, because electron scattering inflates the luminosity signal. Others noted that the Lambda Cold Dark Matter framework — which accurately predicts the structure of the universe at every other scale — couldn’t easily produce so many massive objects so early.

Black holes wearing disguises

Chen and Mo’s model places the little red dots squarely inside standard cosmological physics. In their picture, the dots are supermassive black hole seeds — formed within dense nuclear star clusters at the cores of early galaxies — undergoing what the researchers call “nuclear bursts”: short, violent episodes triggered when two galaxies pass close enough to gravitationally disturb each other’s central gas reservoirs.

During a nuclear burst, gas floods toward the black hole faster than it can be radiated away. The system enters super-Eddington accretion, a regime where the infalling matter forms a thick, optically opaque disk that traps much of the outward radiation and funnels it into narrow polar jets rather than spreading it spherically. The dense surrounding cocoon of gas and dust then absorbs the remaining light and re-emits it in the infrared — producing the characteristic red color and compact appearance that gave the objects their name.

How fast is too fast?

The Eddington limit defines a balance point: above a critical luminosity, the outward pressure of radiation on infalling gas should exceed the inward pull of gravity, shutting off accretion. For a black hole of a million solar masses, this corresponds to a maximum accretion rate of roughly 22 solar masses per year.

The Chen and Mo model requires rates an order of magnitude beyond that. Whether such rates are physically achievable has been debated for decades. Numerical simulations of super-Eddington accretion exist, and observations of ultraluminous X-ray sources in nearby galaxies suggest the regime is real. The JWST little red dots, if this model is correct, would represent the most extreme and numerous population of super-Eddington accretors ever identified — not exotic exceptions, but a widespread early-universe phase that most massive black holes passed through.

A compelling fit, still awaiting peer review

The model is consistent with the distribution of little red dots seen by JWST — their abundance, their clustering at redshifts between 5 and 8 (roughly 1–2 billion years after the Big Bang), and their typical luminosities. It also explains their compactness: the accreting region is small, wrapped in obscuring gas, and luminous enough to outshine any surrounding galaxy.

The caveat matters: this is a preprint, posted to arXiv in May 2026, not yet peer-reviewed. A perspective piece in Science engaged with the model’s implications, but independent verification of its key predictions has not yet been published. The competing electron-scattering explanation and the Chen-Mo nuclear burst model may not be mutually exclusive — some dots could be super-Eddington seeds, others lower-mass objects with amplified apparent brightness.

The seed problem

If confirmed, the implications extend well beyond explaining one class of JWST curiosity. The central puzzle in high-redshift cosmology — how the universe assembled supermassive black holes weighing billions of solar masses within the first billion years — has lacked a satisfying mechanism. Standard growth at or below the Eddington limit cannot build those masses fast enough, even starting from the largest plausible seeds. Episodic super-Eddington bursts, triggered by the frequent galaxy interactions of the crowded early universe, could compress the required growth time into a viable window.

The nuclear burst model predicts unsteady growth: bright phases lasting tens of thousands of years, separated by quieter intervals. JWST cannot watch a single black hole grow across cosmic time, but it can take a statistical census of a population caught at different phases. The coming decade of JWST observations — combined with data from the Roman Space Telescope and next-generation X-ray observatories — will either vindicate the picture or refine it toward something stranger.

Frequently asked questions

What are JWST’s “little red dots”?
Compact, faint, very red objects found in large numbers in JWST deep-field images of the early universe. Their high luminosity for their small apparent size, extreme redness, and concentration in the first billion years after the Big Bang made them a subject of active debate from 2023 onward.

What is the Eddington limit?
A theoretical maximum accretion rate for a black hole, defined by the balance between gravitational pull and outward radiation pressure. The Chen and Mo model proposes that early black holes regularly exceeded this limit by a factor of ten during episodes of rapid galaxy interaction.

Is this paper peer-reviewed?
No. The paper by Yangyao Chen and Houjun Mo appeared on arXiv in May 2026 as a preprint and has not yet undergone formal peer review. A related perspective piece was published in Science, but the model’s key predictions have not been independently confirmed.

Does this solve the “seed problem” of early black holes?
It offers a plausible pathway. If super-Eddington accretion is as common in the early universe as this model predicts, it provides a mechanism for building billion-solar-mass black holes within the first billion years of cosmic history — a timeline that standard growth rates cannot achieve.

Reference: Chen, Y. & Mo, H. J. (2026). Nuclear burst model of the little red dots from JWST: black hole seeds in the early universe. arXiv preprint arXiv:2605.31077. Perspective: Harikane, Y. & Inoue, A. K. (2026). Science, 10.1126/science.adz8603.

Tags: astronomy, James Webb Space Telescope, JWST, black holes, little red dots, super-Eddington accretion