JWST's Little Red Dots May Be Feeding Black Holes Too Fast

Webb's Little Red Dots look like baby black holes in gas cocoons. New models say they gorged in bursts, but X-ray data and revised weights complicate the story.

Little Red Dots are compact, red galaxies that the James Webb Space Telescope keeps finding in the first billion years of the universe, and many astronomers now think each one hides a growing black hole. The "too fast" part is a proposal, not a measurement. A May 2026 preprint by Yangyao Chen and Houjun Mo argues these black holes feed in short bursts at up to roughly 10 times the Eddington limit, the usual speed cap on black hole growth. A September 2026 Nature simulation paints a similar picture. Chandra X-ray data and a Nature spectroscopy study pull the other way, and the dispute hinges on one number: how heavy these black holes really are.

What Are JWST's Little Red Dots?

Little Red Dots (LRDs) are point-like galaxies with a V-shaped spectrum. The slope changes near the Balmer limit, a hydrogen feature at about 365 nanometers, as Rusakov and colleagues describe. Many also show hydrogen emission lines wider than 1,000 kilometers per second. A 2024 survey led by Matthee found them to be an abundant population around redshift 5.

Because of light travel time, an LRD at redshift 5 shows the universe roughly a billion years after the Big Bang. The telescope had already found unexpectedly mature early galaxies, so LRDs arrived in a field that was already short on easy explanations.

What Does "Feeding Too Fast" Mean for a Black Hole?

Gas orbiting a black hole spirals inward and heats up, and the light it gives off pushes back on the gas that follows. The Eddington limit is the brightness where that outward push balances gravity, so it roughly caps how fast a black hole can grow. Super-Eddington accretion means beating the cap, which can work when light gets trapped in the inflow and is dragged inward with the gas.

The simulation by Chon and colleagues puts numbers on it: accretion at several to a few tens of times the Eddington rate, lasting less than a million years. Why bother? Supermassive black holes already exist less than a billion years after the Big Bang, and how they grew so fast is still unclear. Beating the cap is one way to buy time.

What Do the New Papers Actually Claim?

Chen and Mo (preprint, May 2026). In their model, galaxy nuclei go through "nuclear bursts" with super-Eddington feeding at about 10 times the Eddington rate. Young stars born in the burst supply the blue ultraviolet light, and the feeding black hole supplies the red optical glow. Most LRDs at redshift 5 were seeded at redshift 20 or higher, then grew to about 100,000 to 10 million solar masses. The authors call the observed LRDs the "tip of the iceberg" of a larger, fainter population. As of October 2026 I found it only as an arXiv preprint, with no sign of peer review.

Chen and Mo's pitch is that LRDs fall out of standard cold dark matter cosmology without new physics.

Chon and colleagues (Nature, September 16, 2026). Their simulation forms heavy seeds of about a million solar masses in dense protocluster regions flooded with far-ultraviolet light. A brief super-Eddington phase then grows them to roughly 30 million solar masses by redshift 8. In the model, the obscured, LRD-like phase ends after several hundred thousand years, and the object stays X-ray dark for at least half a million. One caveat matters: the model has no jets or winds from the black hole, and the authors call it an upper limit on growth.

Why Do Other Astronomers Doubt the Burst Story?

Andrea Sacchi and Ákos Bogdán stacked Chandra data for 55 LRDs, nearly 400 million seconds of exposure, and still saw no X-rays. They concluded that current super-Eddington models are ruled out unless the gas is extremely thick, and they speculated that the black holes are less massive and less luminous than believed.

Rusakov's team attacks the weight itself. In the best JWST spectra, electron scattering, not fast orbital motion, broadens the lines. Remove that effect and the black hole masses drop by about a factor of 100, to 105 to 107 solar masses, with feeding near the Eddington limit instead of far above it.

Radio telescopes have shown how dust can hide early galaxies, so dust was the obvious suspect for the red color. A 2026 analysis by Hviding and colleagues of an X-ray-detected LRD found that standard dust models could not reproduce its colors without extremely steep extinction curves. A stars-only explanation from Baggen and colleagues argued that very dense star populations could mimic the broad lines, though Rusakov's team says the luminosity in its sample needs accretion.

Do the Two Camps Disagree as Much as They Seem To?

Less than the headlines suggest. Both sides land on black holes of roughly 100,000 to 10 million solar masses. Chen and Mo say super-Eddington bursts built them. Rusakov's team says they now feed near the Eddington limit and were never as heavy as the old estimates implied. That overlap is my reading, not a claim either team makes.

If the lower masses hold, "feeding too fast" turns into a different question: how did black holes get started so early? That is the seed problem Chon's simulation tackles with million-solar-mass starting points. Until the weights are settled, the "too fast" label depends on which scale you trust.

What Would Settle the Question?

Rusakov's team says larger samples and repeat observations are needed to test how well current mass methods work for these objects. Deeper X-ray stacks would test Sacchi and Bogdán's result. Peer review would test Chen and Mo.

Pulsar timing arrays already hint at a gravitational-wave background likely produced by merging supermassive black holes. Chon's team says heavy-seed scenarios predict tens to hundreds of LISA events over a three-year mission, which would be a direct test.

Conclusion

Little Red Dots may well be baby black holes wrapped in gas. Whether they fed too fast depends on a weight nobody can measure directly yet.

If the lower masses hold, the puzzle shifts from overeating to how black holes got started so early. For now, "too fast" is a hypothesis with a simulation behind it and an X-ray non-detection against it.

Frequently Asked Questions

Are Little Red Dots black holes or star clusters?

Most of the evidence now favors accreting black holes inside dense gas, but a stars-only picture has not been fully excluded. The answer may differ from object to object.

How big are the black holes in Little Red Dots?

Rusakov and colleagues estimate 100,000 to 10 million solar masses, about 100 times lower than earlier estimates. Earlier work argued for 10 million to 100 million.

Why are Little Red Dots red?

One leading explanation is dense ionized gas around the black hole, not ordinary dust. That gas reprocesses the light into the red optical glow and the V-shaped spectrum.

Do Little Red Dots still exist today?

A few similar systems have been reported at later cosmic times, but they appear to be exceptionally rare. The population is mainly an early-universe phenomenon.

Has the Chen and Mo paper been peer reviewed?

As of October 2026 I found it listed only as an arXiv preprint. Treat its numbers as a proposal until a journal version appears.