A team of astronomers led by Professor Linhua Jiang at the Kavli Institute for Astronomy and Astrophysics and the School of Physics at Peking University has discovered two unusual little red dots that appear to be in transition toward typical active galactic nuclei or quasars. This finding provides direct observational evidence for the evolutionary fate of little red dots and suggests that at least some of them represent the early, short-lived phase of rapidly growing supermassive black holes. The result was published in Nature Astronomy on September 16, 2026 (volume 10, pages 1372–1386), as an article entitled “The discovery of two little red dots in transition into quasars”. Graduate students Shuqi Fu and Zijian Zhang are the co-first authors of the paper.
Quasars are among the brightest and most energetic objects in the Universe. Their power comes from supermassive black holes that are actively accreting gas at the centers of galaxies. As gas falls into the region around a black hole, it releases intense radiation that can be seen across billions of light years. Quasars illuminate the distant Universe and strongly affect the formation and evolution of their host galaxies.
A key question is how these supermassive black holes grew so rapidly in the early Universe. In recent years, the James Webb Space Telescope has uncovered a new population of compact red sources known as “little red dots” (LRDs). These objects show a distinctive V-shaped spectral energy distribution: a blue ultraviolet continuum and a red optical continuum. Many of them also show broad emission lines, indicating the possible presence of rapidly accreting supermassive black holes.
Yet most LRDs lack the strong X-ray, radio, and mid-infrared emission commonly seen in typical active galactic nuclei. They seem to contain growing black holes, but those black holes are not fully revealed. The nature of LRDs, and whether they evolve into ordinary active galactic nuclei or quasars, have therefore remained open questions.
To address this problem, Jiang’s team constructed a parent sample of LRDs using JWST/NIRCam multi-band imaging and optical morphology. They then cross-matched the sample with the Chandra COSMOS X-ray catalog and identified two exceptional LRDs, named Forge I and Forge II.
Figure 1. An artist’s conception of little red dots in transition into quasars (credit: Jingchuan Yu, Beijing Planetarium; Shuqi Fu, Zijian Zhang, Linhua Jiang, Peking University).
The two objects lie at redshifts of 2.871 and 2.930, meaning that we see them as they were about 11 billion years ago. They satisfy the defining criteria of LRDs: a clear V-shaped spectral energy distribution (Figure 2a), a compact optical morphology (Figure 2b), and broad emission lines (Figure 2c). Unlike normal LRDs, however, Forge I and Forge II also exhibit intense X-ray, radio, and mid-infrared radiation, which are hallmark signatures of typical active galactic nuclei and quasars.
Figure 2. Multi-wavelength properties of Forge I and Forge II. (a) Spectral energy distributions compared with those of other LRDs. (b) Compact optical morphology revealed by JWST/NIRCam F444W images. (c) Broad emission lines and strong absorption features shown in the JWST/NIRCam F444W slitless spectra. (d) Location of Forge I and Forge II on the Lbol-MBH plane, between LRDs and quasars.
The team further studied the two sources using JWST/NIRCam F444W slitless spectroscopy. The spectra reveal broad He I, Paγ, and O I emission lines, with widths of several thousand kilometers per second, confirming their active galactic nucleus nature (Figure 2c). Using the broad Paschen emission lines and the 1 micron continuum luminosity, the team estimated black hole masses of about 370 million and 820 million solar masses for Forge I and Forge II, respectively. These masses are already comparable to those of quasars (Figure 2d).
The spectra also reveal complex gas motions. Strong He I absorption lines may indicate inflowing or fallback gas near the central black holes. Forge II also shows tentative blueshifted Paδ absorption (Figure 2c), suggesting that part of the gas may be moving outward. These features point to a rapidly changing environment around the growing black holes.
Multi-band image decomposition and spectral energy distribution modeling provide a coherent picture of this transition. In the ultraviolet, the observed light is still dominated by the host galaxy or extended nebular emission. At optical to near-infrared wavelengths, however, the central point source becomes dominant. The strong mid-infrared emission likely traces an emerging hot-dust torus.
In this scenario, typical LRDs are rapidly growing black holes enshrouded by dense gas. The dense envelope absorbs or blocks much of the high-energy radiation, preventing strong X-ray, radio, and mid-infrared signals from escaping. Forge I and Forge II appear to be caught as this envelope is dispersing. As the central black hole grows, the surrounding gas becomes porous or is cleared away, allowing X-ray and radio emission to escape while a hot-dust structure begins to form.
In the COSMOS field, the team found seven broad-line LRDs at redshift around 3. Only Forge I and Forge II show the transitional properties reported here. By comparing the number density of the Forges with that of all LRDs, the team estimated that the transitional phase lasts for only about 6% of the LRD lifetime. This short timescale helps explain why such objects have been difficult to catch before.
In summary, this work identifies two LRDs that are transitioning into quasars and provides a key link between LRDs and normal active galactic nuclei. It suggests that at least some LRDs are not isolated mysteries, but early stages in the growth of supermassive black holes. As the dense gas cocoon around the black hole disperses, high-energy radiation escapes and the system gradually reveals its quasar-like nature.
This work was supported by the National Science Foundation of China and other funding agencies.
Link to the paper:
https://www.nature.com/articles/s41550-026-02885-8
