GS-z14: Pioneering High-z Galaxy Study

• GS-z14 is a designation for high-redshift JADES galaxies in GOODS-South, notably GS-z14-0 confirmed at z~14 by its Lyman-break and rest-frame optical signatures.
• Spectroscopic and far-infrared observations, including JWST/NIRSpec and ALMA measurements, precisely determine its redshift, luminosity, star formation rate, and metallicity.
• The detailed physical and morphological characterization of GS-z14 provides a critical benchmark for understanding early galaxy formation and calibrating cosmic dawn models.

GS-z14 designates the JADES galaxies whose shorthand names begin with JADES-GS-z14 in the GOODS-South field. In the discovery paper, the term covered both JADES-GS-z14-0 and JADES-GS-z14-1, but in later informal usage it usually denotes JADES-GS-z14-0, the brighter, more extended source. The system became central to early-galaxy studies because JWST/NIRSpec established a Lyman-break redshift near $z\sim14$, ALMA later fixed the systemic redshift of GS-z14-0 at $z=14.1793\pm0.0007$ or $z=14.1796\pm0.0007$, and subsequent JWST/MIRI spectroscopy opened direct rest-frame optical diagnostics only $\sim 300$ Myr after the Big Bang (Carniani et al., 2024, Schouws et al., 2024, Helton et al., 22 Dec 2025).

1. Nomenclature, field placement, and basic identification

In JADES nomenclature, JADES-GS-z14-0 is the extended source JADES-GS-53.08294-27.85563 at RA, Dec (ICRS) = 03:32:19.905, $-27{:}51{:}20.27$, with NIRCam catalog ID 183348, while JADES-GS-z14-1 is JADES-GS-53.07427-27.88592. GS-z14-0 lies only 0.4″ east of a foreground galaxy at $z=3.475$, a geometry that initially raised the possibility of a low-$z$ interloper; later spectroscopy ruled that out (Carniani et al., 2024).

Source	Initial spectroscopic estimate	Basic characterization
JADES-GS-z14-0	$z=14.32^{+0.08}_{-0.20}$	Brighter, more extended, higher-$z$ source
JADES-GS-z14-1	$z=13.90\pm0.17$	Fainter, more compact source

GS-z14-0 is only mildly magnified, with total lensing magnification $z=14.1793\pm0.0007$0 from nearby foreground galaxies, so its luminosity is essentially intrinsic rather than lens-amplified (Carniani et al., 2024). GS-z14-1 is separated by 1.9′ on the sky from GS-z14-0, corresponding to $z=14.1793\pm0.0007$1 comoving Mpc at $z=14.1793\pm0.0007$2; whether the two belong to a common large-scale structure remains unresolved because narrow-line redshifts are still limited (Carniani et al., 2024).

2. Spectroscopic establishment of the $z=14.1793\pm0.0007$3 nature

The decisive observational signature in the original NIRSpec prism data was a sharp continuum break with no detectable flux blueward of $z=14.1793\pm0.0007$4. The break was modeled as a rest-UV power law,

$z=14.1793\pm0.0007$5

combined with IGM absorption, possible damped Ly$z=14.1793\pm0.0007$6 absorption, and optional Ly$z=14.1793\pm0.0007$7 emission. This yielded $z=14.1793\pm0.0007$8 for GS-z14-0 and $z=14.1793\pm0.0007$9 for GS-z14-1 (Carniani et al., 2024).

For GS-z14-0, the prism spectrum also showed a tentative C III] $z=14.1796\pm0.0007$0 feature at $z=14.1796\pm0.0007$1 with 3.6$z=14.1796\pm0.0007$2 significance, implying

$z=14.1796\pm0.0007$3

and

$z=14.1796\pm0.0007$4

If that line is adopted as systemic, the Ly$z=14.1796\pm0.0007$5 break requires a damped absorber with

$z=14.1796\pm0.0007$6

No comparably secure UV line was detected in GS-z14-1 (Carniani et al., 2024).

ALMA subsequently detected [O III] $z=14.1796\pm0.0007$7 from GS-z14-0 at 223.524 GHz with 6.6–6.67$z=14.1796\pm0.0007$8 significance, fixing the redshift at $z=14.1796\pm0.0007$9 or $\sim 300$0 and improving the redshift precision by $\sim 300$1 relative to the break-only estimate (Schouws et al., 2024, Carniani et al., 2024). Ultra-deep MIRI/LRS observations later detected rest-frame optical [O III] $\sim 300$2 at $\sim 300$3 and H$\sim 300$4 at $\sim 300$5, with an [O III]-based redshift $\sim 300$6 consistent with the ALMA result (Helton et al., 22 Dec 2025).

A later re-analysis of the damping wing at fixed systemic redshift $\sim 300$7 found

$\sim 300$8

and showed that an IGM + DLA model is strongly preferred over an IGM-only damping wing, with $\sim 300$9 (Heintz et al., 9 Feb 2025). This shifted the interpretation from an uncertain break placement to a well-defined systemic redshift plus substantial neutral absorption local to the source.

3. Photometry, luminosity, and structural classification

For GS-z14-0, NIRCam and MIRI photometry show a classic dropout pattern blueward of the break and bright detections redward of it. In the original JADES analysis the key fluxes included F182M $-27{:}51{:}20.27$0 nJy, F200W $-27{:}51{:}20.27$1 nJy, and F444W $-27{:}51{:}20.27$2 nJy, while MIRI gave F770W $-27{:}51{:}20.27$3 nJy (Carniani et al., 2024). A dedicated MIRI analysis measured the same source at $-27{:}51{:}20.27$4 nJy in F770W and found a flux excess relative to F444W of

$-27{:}51{:}20.27$5

with the most plausible interpretation that at least one-third of the $-27{:}51{:}20.27$6 flux comes from H$-27{:}51{:}20.27$7 and/or [O III] $-27{:}51{:}20.27$8 (Helton et al., 2024).

Its intrinsic UV luminosity is

$-27{:}51{:}20.27$9

after correcting for $z=3.475$0, placing it among the brightest known galaxies at $z=3.475$1 (Carniani et al., 2024). In the photometric-detection paper, the most plausible SED solution contained half a billion solar masses in stars together with a strong burst of star formation in the most recent few Myr (Helton et al., 2024).

Morphologically, GS-z14-0 is resolved in NIRCam. Radial profiles in F277W extend to $z=3.475$2 kpc and are broader than a field star, GN-z11, and GHZ2. ForcePho modeling yields a single-component Sérsic profile with $z=3.475$3 and a deconvolved UV half-light radius

$z=3.475$4

using $z=3.475$5 kpc at $z=3.475$6 (Carniani et al., 2024). This size later became a defining element in classifying GS-z14-0 as an extended, luminous, weak-line galaxy rather than a compact extreme-line emitter. PAN-z14-1 at $z=3.475$7, with $z=3.475$8 pc, $z=3.475$9, and weak UV lines, was explicitly described as “remarkably similar” to GS-z14-0 (Donnan et al., 16 Jan 2026).

GS-z14-1 is materially different. In the initial confirmation it was only marginally resolved and consistent with

$z$0

(Carniani et al., 2024). A later ultra-deep study refined this to an unresolved source across 16 NIRCam bands, with

$z$1

and

$z$2

making it the faintest spectroscopically confirmed galaxy at $z$3 and one of the smallest known systems in that regime (Wu et al., 30 Jul 2025).

4. Stellar populations, rest-frame optical spectroscopy, and ionizing output

Early SED fits already suggested that GS-z14-0 is dominated by stellar continuum rather than nebular continuum or AGN light. The original BEAGLE analysis, fitting NIRSpec plus NIRCam and MIRI photometry, gave

$z$4

$z$5

$z$6

$z$7

and a favored high-escape-fraction solution,

$z$8

because that scenario most naturally reproduced both the blue UV slope and the absence of strong UV lines (Carniani et al., 2024). A Prospector analysis anchored by ALMA [O III] instead found $z$9, a non-zero escape fraction of ionizing photons of $z=14.32^{+0.08}_{-0.20}$0, $z=14.32^{+0.08}_{-0.20}$1, $z=14.32^{+0.08}_{-0.20}$2, and a mass-weighted age of $z=14.32^{+0.08}_{-0.20}$3 Myr (Carniani et al., 2024). The numerical differences are model-dependent, but both approaches imply a rapidly assembled, metal-enriched, still-young stellar system.

Ultra-deep MIRI/LRS spectroscopy materially sharpened the rest-frame optical picture. It detected [O III] $z=14.32^{+0.08}_{-0.20}$4 at $z=14.32^{+0.08}_{-0.20}$5 and H$z=14.32^{+0.08}_{-0.20}$6 at $z=14.32^{+0.08}_{-0.20}$7, with measured fluxes

$z=14.32^{+0.08}_{-0.20}$8

and

$z=14.32^{+0.08}_{-0.20}$9

From H$z$0, the inferred recent star-formation rate is

$z$1

and, using $z$2 pc,

$z$3

The same spectrum implies

$z$4

or $z$5 in the Cue photoionization fit (Helton et al., 22 Dec 2025). That value is high but not extreme: it is consistent with canonical reionization-era assumptions rather than requiring a radically top-heavy IMF.

The same MIRI study found a substantial metallicity ambiguity. Standard strong-line diagnostics gave

$z$6

$z$7

$z$8

and

$z$9

whereas detailed Cue photoionization modeling instead yielded

$z=13.90\pm0.17$0

and

$z=13.90\pm0.17$1

The paper explicitly attributes this discrepancy to strong density-metallicity degeneracies and to the limited set of rest-frame optical constraints presently available (Helton et al., 22 Dec 2025).

GS-z14-1 occupies a lower-mass, fainter part of the same redshift frontier. Its deep Prospector fit gives

$z=13.90\pm0.17$2

$z=13.90\pm0.17$3

and a gas-phase metallicity posterior centered near $z=13.90\pm0.17$4 with a sharp drop above $z=13.90\pm0.17$5 (Wu et al., 30 Jul 2025). Its tentative MIRI/F770W detection of $z=13.90\pm0.17$6 nJy implies

$z=13.90\pm0.17$7

weaker than in most luminous $z=13.90\pm0.17$8 galaxies (Wu et al., 30 Jul 2025).

5. Neutral gas, far-infrared lines, dust, and the disputed gas budget

GS-z14-0 is one of the few $z=13.90\pm0.17$9 galaxies with both rest-UV and far-infrared diagnostics. ALMA detected

$z=14.1793\pm0.0007$00

or equivalently $z=14.1793\pm0.0007$01, while deep [C II] observations found only

$z=14.1793\pm0.0007$02

implying

$z=14.1793\pm0.0007$03

depending on the adopted [O III] measurement (Schouws et al., 2024, Schouws et al., 3 Feb 2025). In Cloudy-based analyses, this ratio requires a highly ionized ISM, with $z=14.1793\pm0.0007$04 in the [C II] non-detection paper and $z=14.1793\pm0.0007$05 to $z=14.1793\pm0.0007$06 in earlier ALMA+MIRI modeling (Schouws et al., 3 Feb 2025, Schouws et al., 2024). The [C II] study further inferred a relatively low gas density,

$z=14.1793\pm0.0007$07

a gas-phase metallicity of

$z=14.1793\pm0.0007$08

and a molecular-gas-based upper limit

$z=14.1793\pm0.0007$09

arguing for strong feedback, low gas content, and a rapidly evolving ISM (Schouws et al., 3 Feb 2025).

A different framework, however, followed from the DLA interpretation of the Ly$z=14.1793\pm0.0007$10 break. Combining the NIRSpec damping wing with ALMA [O III] and [C II] information, Heintz et al. inferred a total gas mass

$z=14.1793\pm0.0007$11

a gas fraction

$z=14.1793\pm0.0007$12

and a picture in which the bulk of the baryons are in a massive, largely pristine neutral H I reservoir surrounding a compact star-forming core (Heintz et al., 9 Feb 2025). In that analysis, the DLA line-of-sight dust-to-gas ratio is

$z=14.1793\pm0.0007$13

and the DLA metallicity implied by that ratio is in the range

$z=14.1793\pm0.0007$14

substantially lower than the metallicity of the central H II region (Heintz et al., 9 Feb 2025). The central tension is therefore geometrical: [O III]-based dynamics and [C II]-based gas proxies emphasize the compact core, whereas the damping-wing analysis emphasizes a much larger H I structure.

Dust constraints are consistently stringent, though method-dependent. ALMA continuum non-detections imply

$z=14.1793\pm0.0007$15

(Heintz et al., 9 Feb 2025), and the 2024 ALMA [O III] paper quoted a dust-to-stellar-mass ratio

$z=14.1793\pm0.0007$16

(Schouws et al., 2024). An $z=14.1793\pm0.0007$17-based estimate assuming dust distributed over $z=14.1793\pm0.0007$18 instead gives

$z=14.1793\pm0.0007$19

for GS-z14-0, with a dust-to-stellar-mass ratio $z=14.1793\pm0.0007$20 (Carniani et al., 2024). Those numbers should not be combined as if they were identical measurements; they instead bracket a shared conclusion that GS-z14 is dust-poor relative to many lower-redshift galaxies.

6. Competing interpretations and comparison objects

The dominant interpretation in the discovery and follow-up literature is that GS-z14-0 is stellar-continuum dominated. The decisive arguments are morphological extension, the absence of strong high-ionization UV lines such as N V, C IV, and He II, and UV slopes consistent with young stellar populations. The original JADES paper therefore concluded that the excess of luminous galaxies at cosmic dawn cannot be explained simply by AGN accretion light (Carniani et al., 2024).

Comparison with PAN-z14-1 reinforced that view. PAN-z14-1, spectroscopically confirmed at $z=14.1793\pm0.0007$21, has $z=14.1793\pm0.0007$22, $z=14.1793\pm0.0007$23, $z=14.1793\pm0.0007$24 pc, and no rest-UV lines at $z=14.1793\pm0.0007$25. It was explicitly described as “remarkably similar” to GS-z14-0 and used to support an emerging class of large, luminous, weak-line galaxies at $z=14.1793\pm0.0007$26, distinct from compact strong-line emitters such as GN-z11, GHZ2, and MoM-z14 (Donnan et al., 16 Jan 2026).

Alternative interpretations have nevertheless been proposed. One accretion-disc paper argued that JADES-GS-z14-1, not the usual GS-z14-0, has a UV slope compatible with a standard thin disc and inferred a minimum sub-Eddington black-hole mass

$z=14.1793\pm0.0007$27

with

$z=14.1793\pm0.0007$28

if the UV continuum were disc-dominated (Fabian et al., 5 Sep 2025). The same paper did not argue that GS-z14-0 itself is AGN-dominated, and the broader literature still treats GS-z14-0 as primarily stellar.

A more speculative proposal identified both GS-z14-0 and GS-z14-1 as supermassive Dark Star candidates, with best-fit Dark Star masses of

$z=14.1793\pm0.0007$29

for GS-z14-0 and

$z=14.1793\pm0.0007$30

for GS-z14-1 (Ilie et al., 9 May 2025). That work highlighted a tentative He II $z=14.1793\pm0.0007$31 absorption feature in GS-z14-0 with $z=14.1793\pm0.0007$32 and rest-frame $z=14.1793\pm0.0007$33 Å, but also noted that the ALMA [O III] detection makes a simple isolated Dark Star interpretation unlikely because it implies a metal-enriched environment (Ilie et al., 9 May 2025). At present, these alternatives remain minority interpretations rather than the working consensus.

7. Role in early-galaxy theory and large-scale structure

GS-z14 rapidly became a benchmark for simulation studies. A FIRE/IllustrisTNG post-processing analysis found no direct analog to GS-z14 or GHZ2 at $z=14.1793\pm0.0007$34; the closest FIRE analog, z5m12b, matches GS-z14-like [O III] luminosity, SFR, stellar mass, and line width only at $z=14.1793\pm0.0007$35, lagging GS-z14 by roughly 300 Myr (Yang et al., 25 Apr 2025). The paper argued that limited simulation volume and resolution could explain part of the discrepancy, but not the object’s extremity.

Protocluster modeling in FOREVER22 reached a more environmental conclusion. Using overdensity $z=14.1793\pm0.0007$36 and the stellar mass of the most massive member as protocluster diagnostics, it argued that GS-z14-0 and GS-z14-1 are likely progenitors of Coma-type clusters, because GS-z14-0 lies well above the redshift-dependent critical stellar mass for a Coma progenitor and because massive galaxies at such early times preferentially occupy overdense peaks in the simulations (Morokuma-Matsui et al., 9 Apr 2025).

A later Renaissance-based assessment sharpened the model tension. It treated GS-z14 with

$z=14.1793\pm0.0007$37

at

$z=14.1793\pm0.0007$38

and concluded that, unlike MoM-z14, GS-z14 remains an outlier even after allowing for generous cosmic-variance boosts. The authors interpreted the mismatch as primarily astrophysical rather than cosmological: current prescriptions appear to underproduce star formation in the most massive and earliest halos rather than falsifying $z=14.1793\pm0.0007$39CDM outright (McCaffrey et al., 9 Sep 2025).

The empirical significance is therefore broader than a single record redshift. GS-z14 links several previously separate issues: the unexpectedly slow decline of the bright end of the UV luminosity function, the presence of chemically enriched and line-emitting gas by $z=14.1793\pm0.0007$40, the existence of both extended weak-line and compact weak-line systems at the same epoch, and the need for simulations to produce high stellar masses and strong bursts on $z=14.1793\pm0.0007$41 Myr timescales (Carniani et al., 2024, Helton et al., 22 Dec 2025). A plausible implication is that early galaxy formation is not adequately described by a single mode: GS-z14-0, PAN-z14-1, GS-z14-1, GHZ2, GN-z11, and MoM-z14 collectively suggest a heterogeneous population spanning extended stellar systems, ultracompact starbursts, and possibly very different gas geometries (Donnan et al., 16 Jan 2026, Wu et al., 30 Jul 2025).

The principal unresolved questions are now precise rather than generic: whether GS-z14-0’s metallicity is closer to $z=14.1793\pm0.0007$42 or $z=14.1793\pm0.0007$43; whether its baryons are centrally gas-poor or globally H I-dominated; how large its LyC escape fraction actually is; and whether GS-z14-0 and GS-z14-1 inhabit a common overdense structure. Those issues define its continuing value as a calibration object for JWST, ALMA, and high-redshift galaxy-formation theory.