JWST data reveal candidate broad-line AGN at z ≈ 10, offering tentative virial black hole mass

A new arXiv preprint says public James Webb Space Telescope data contain what its authors identify as the highest-redshift broad-line active galactic nucleus yet reported, at z = 9.9837 — a result that, if it holds up, would enable the first virial black hole mass estimate at roughly z = 10.

The object, called MoM-BL-z10, is described in a paper posted to arXiv on Oct. 20 by lead author Xingyao Cai of Waseda University and colleagues. The study is a preprint, meaning it has not been peer reviewed. The team says it found the source in a systematic archival search of public JWST/NIRSpec PRISM spectra in the DAWN JWST Archive, drawing from 1,247 galaxies and galaxy-like sources with robust spectroscopic redshifts between 6.0 and 10.3.

The central evidence is a reported broad H-beta emission line at about 5.35 microns. Broad-line active galactic nuclei, or AGN, are identified by very wide emission lines produced by gas moving at thousands of kilometers per second close to a black hole. In this case, the authors report the broad H-beta feature at a signal-to-noise ratio of 3.87. Their preferred fit, which includes both narrow and broad components, gives an intrinsic line width of about 3,480 kilometers per second, though with large uncertainties. The paper also points to other AGN-like signs, including detection of neon-III emission and line ratios such as high [Ne III]/[O II] and [O III]/H-beta. In imaging, the source appears compact, and the authors say its continuum shape and tentative mid-infrared detection make it resemble a so-called Little Red Dot source.

That broad line is what makes the claim especially notable. “The broad Hβ detection enables the first virial SMBH mass estimate at z ∼ 10,” Cai and colleagues wrote. Using a standard single-epoch virial method — which combines line width and brightness to estimate black hole mass — the team derives a mass of 3.6 × 10^7 times the mass of the sun and an Eddington ratio, a measure of accretion relative to a theoretical limit, of 0.11. They also present a more model-dependent scenario tailored to Little Red Dot-style assumptions: if the source is accreting at the Eddington limit and a different bolometric correction is used, the black hole mass would be 2.1 × 10^6 solar masses. In the authors’ cosmology, z = 9.98 corresponds to a universe about 470 million years old, placing a putative million- to tens-of-millions-solar-mass black hole extremely early in cosmic history.

The authors say that would make MoM-BL-z10 “a new observational benchmark for understanding the early growth of supermassive black holes.” More broadly, JWST is now reaching much fainter and earlier black hole candidates than astronomers could access before; prior to JWST, the best-known confirmed quasars were generally found only out to about z = 7.7 in ground-based surveys.

Still, the evidence comes with important caveats. The reported broad H-beta feature is modest in significance and falls near the long-wavelength edge of NIRSpec/PRISM, where instrumental systematics can be more problematic. The authors themselves say confirmation with higher-quality spectroscopy would be valuable. And the black hole mass estimate is not a direct weighing: It relies on a virial method calibrated at lower redshift, while the inferred mass shifts substantially depending on assumptions about bolometric correction and how fast the object is accreting.

If confirmed, the object would surpass the previous broad-line benchmark cited in the paper, CAPERS-LRD-z9 at z = 9.288, published in 2025. The distinction matters. Cai and colleagues are claiming a record for a broad-line AGN — not for every possible AGN candidate at similar redshift. The paper notes that other candidate AGN may exist near z ~ 10, but says they do not yet show a securely detected broad optical emission line of the kind needed for this type of black hole mass estimate.