AT2024wpp 'Whippet': A Cow-like LFBOT

• AT2024wpp is a luminous fast blue optical transient marked by a rapid ~4.3-day rise to a superluminous peak and a persistently hot, nearly featureless spectrum.
• Observations indicate near-relativistic outflow speeds, strong radio and X-ray emission, and minimal polarization, supporting an engine-driven explosion with constrained ejecta geometry.
• Multiwavelength data, from optical to radio and X-rays, suggest models invoking black-hole accretion and dense circumstellar environments to explain its unusual photometric and spectral evolution.

AT2024wpp, also nicknamed “Whippet,” is a luminous fast blue optical transient (LFBOT) of the AT2018cow-like, or “Cow-like,” subclass. It is distinguished by an extremely rapid rise, very high UV–optical luminosity, persistently hot and nearly featureless spectra, strong radio and X-ray emission, and unusually constraining optical polarimetry. Observational studies identify it as one of the most extreme members of the class, while subsequent modeling papers treat it as a critical test case for engine-driven outflows, black-hole accretion scenarios, and the geometry of fast ejecta in Cow-like transients (Pursiainen et al., 2024, Perley et al., 6 Jan 2026, Omand et al., 6 Jan 2026).

1. Discovery, host galaxy, and class membership

AT2024wpp was first identified in ZTF survey data on 2024 September 26, with a first detection reported at MJD 6(0578.4365) and a very rapid brightening over the following day (Pursiainen et al., 2024, Perley et al., 6 Jan 2026). Follow-up spectroscopy tied the transient to a diffuse, extended, star-forming host galaxy at redshift $z = 0.0868$. Different analyses quote distance moduli $\mu \simeq 38.06\,\mathrm{mag}$ and $\mu \approx 37.98\,\mathrm{mag}$ (Pursiainen et al., 2024, Ofek et al., 25 Aug 2025). The source lies at a projected offset of about $5.2\,\mathrm{kpc}$ from the host center, and later high-resolution imaging described the host as a faint, diffuse, face-on disk galaxy with the transient located in the outer region of the stellar disk (Pursiainen et al., 2024, Perley et al., 6 Jan 2026).

The host environment is consistent with the low-mass, moderately star-forming, subsolar-metallicity galaxies often associated with LFBOTs. Prospector modeling gave $\log_{10}(M_*/M_\odot)=8.76\pm0.03$, a stellar age of $\sim1.7\pm0.3\,\mathrm{Gyr}$, and $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$, while KCWI spectroscopy yielded $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$ and $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$ (Perley et al., 6 Jan 2026). Local extinction appears low: no host Na I D absorption was detected in the optical work, and later analyses quoted $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$ toward the transient (Pursiainen et al., 2024, Perley et al., 6 Jan 2026).

Its classification as a Cow-like transient rests on a convergent set of properties: rapid optical evolution, peak luminosity in the superluminous regime, a persistently hot continuum, strong radio and X-ray emission, and close phenomenological similarity to AT2018cow and related events such as AT2022tsd and AT2020xnd (Pursiainen et al., 2024, Ofek et al., 25 Aug 2025). This places AT2024wpp within the small but growing LFBOT population, while also marking it as one of the most luminous and best observed examples (Perley et al., 6 Jan 2026, Omand et al., 6 Jan 2026).

2. Photometric and bolometric evolution

Optically, AT2024wpp was both exceptionally bright and exceptionally fast. Using ATLAS $\mu \simeq 38.06\,\mathrm{mag}$0-band and GOTO $\mu \simeq 38.06\,\mathrm{mag}$1-band photometry, the rest-frame rise to peak was measured as $\mu \simeq 38.06\,\mathrm{mag}$2, with peak apparent magnitude $\mu \simeq 38.06\,\mathrm{mag}$3 and peak absolute magnitude $\mu \simeq 38.06\,\mathrm{mag}$4. The post-peak decline rate at $\mu \simeq 38.06\,\mathrm{mag}$5 d was $\mu \simeq 38.06\,\mathrm{mag}$6, and the total time above half-maximum brightness was $\mu \simeq 38.06\,\mathrm{mag}$7 (Pursiainen et al., 2024). A later UV-through-optical campaign, enabled by pre-peak discovery, derived a peak bolometric luminosity of $\mu \simeq 38.06\,\mathrm{mag}$8, a maximum photospheric radius of $\mu \simeq 38.06\,\mathrm{mag}$9, and a total radiated energy of $\mu \approx 37.98\,\mathrm{mag}$0 (Perley et al., 6 Jan 2026).

The spectral energy distribution was well described by blackbody fits over much of its observed evolution. Early optical and UV data indicated a very blue, hot continuum with a compact radius at $\mu \approx 37.98\,\mathrm{mag}$1 d, implying that the photosphere was still expanding on the way to peak. Throughout the interval from $\mu \approx 37.98\,\mathrm{mag}$2 to $\mu \approx 37.98\,\mathrm{mag}$3 d, the transient remained persistently hot, with $\mu \approx 37.98\,\mathrm{mag}$4, while the inferred radius peaked near maximum light and then receded rapidly (Pursiainen et al., 2024). The basic bolometric relation used in these analyses was

$\mu \approx 37.98\,\mathrm{mag}$5

The photospheric evolution appears atypical for ordinary supernova ejecta. One study inferred early expansion speeds of $\mu \approx 37.98\,\mathrm{mag}$6 up to $\mu \approx 37.98\,\mathrm{mag}$7 d and $\mu \approx 37.98\,\mathrm{mag}$8 between 2.1 and 3.5 d, followed by monotonic contraction of the inferred photospheric radius after $\mu \approx 37.98\,\mathrm{mag}$9–5 d (Perley et al., 6 Jan 2026). Another analysis found that the post-peak radius evolution was approximately linear in time, corresponding to an inward-moving photosphere with effective speed $5.2\,\mathrm{kpc}$0 (Pursiainen et al., 2024). The modeling study parameterized this behavior as $5.2\,\mathrm{kpc}$1 on the rise and $5.2\,\mathrm{kpc}$2 on the decline, while $5.2\,\mathrm{kpc}$3 during the rise and $5.2\,\mathrm{kpc}$4 during the decline (Omand et al., 6 Jan 2026).

The light-curve decline was also unusually steep. After $5.2\,\mathrm{kpc}$5 d, the UV–optical fluxes were described as $5.2\,\mathrm{kpc}$6 in the UV and $5.2\,\mathrm{kpc}$7 in the optical, while the bolometric luminosity decayed approximately as $5.2\,\mathrm{kpc}$8 (Perley et al., 6 Jan 2026). A plausible implication is that the luminosity evolution was governed less by canonical diffusion in a homologously expanding envelope than by a changing reprocessing photosphere in a fast, engine-driven outflow.

3. Spectral evolution from the optical to the far ultraviolet

The optical spectral sequence obtained from $5.2\,\mathrm{kpc}$9 to $\log_{10}(M_*/M_\odot)=8.76\pm0.03$0 rest-frame days showed blue, largely featureless spectra consistent with a hot blackbody continuum at $\log_{10}(M_*/M_\odot)=8.76\pm0.03$1. No narrow transient-associated emission or absorption lines were seen, and overplotted $\log_{10}(M_*/M_\odot)=8.76\pm0.03$2 blackbodies reproduced the continuum shape well (Pursiainen et al., 2024). This persistent featurelessness distinguishes AT2024wpp from interacting supernova subclasses such as Type Ibn and Icn, which typically show stronger circumstellar line signatures.

A tentative broad emission feature appeared near rest-frame $\log_{10}(M_*/M_\odot)=8.76\pm0.03$3. It was broad and shallow, with an implied velocity scale $\log_{10}(M_*/M_\odot)=8.76\pm0.03$4, and strengthened as the light curve declined. The preferred identification was He I $\log_{10}(M_*/M_\odot)=8.76\pm0.03$5, broadened and blueshifted by $\log_{10}(M_*/M_\odot)=8.76\pm0.03$6, whereas an H$\log_{10}(M_*/M_\odot)=8.76\pm0.03$7 identification would require a larger blueshift of order $\log_{10}(M_*/M_\odot)=8.76\pm0.03$8 and was considered less plausible (Pursiainen et al., 2024). This suggests that the optical photosphere at those phases lay inside, or very near, the near-relativistic outflow.

A later spectroscopic study added two further elements. First, optical spectra at days $\log_{10}(M_*/M_\odot)=8.76\pm0.03$9–13 developed very broad depressions near rest-frame $\sim1.7\pm0.3\,\mathrm{Gyr}$0 and $\sim1.7\pm0.3\,\mathrm{Gyr}$1, with characteristic widths corresponding to velocities of order $\sim1.7\pm0.3\,\mathrm{Gyr}$2. These were noted to be reminiscent in shape of Type Ic-BL features, but unlike standard supernova features they faded rather than strengthened (Perley et al., 6 Jan 2026). Second, a combined HST/COS, HST/STIS, and ground-based spectrum at $\sim1.7\pm0.3\,\mathrm{Gyr}$3–21 d showed a single-temperature blackbody continuum with $\sim1.7\pm0.3\,\mathrm{Gyr}$4 across $\sim1.7\pm0.3\,\mathrm{Gyr}$5–$\sim1.7\pm0.3\,\mathrm{Gyr}$6, with deviations $\sim1.7\pm0.3\,\mathrm{Gyr}$7 and no broad intrinsic UV resonance lines. Only narrow ISM-like absorption features at the host redshift and from the Milky Way were detected, together with a damped Ly$\sim1.7\pm0.3\,\mathrm{Gyr}$8 profile implying $\sim1.7\pm0.3\,\mathrm{Gyr}$9 (Perley et al., 6 Jan 2026). In that interpretation, intense ionization in the ejecta and surrounding circumstellar material suppressed normal UV line formation.

At later times, the spectrum changed qualitatively. From 36 to 61 rest-frame days, H$$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$0, possibly H$$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$1, He I $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$2, and He II $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$3 emerged. H$$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$4 displayed a double-peaked profile with one component at systemic velocity and another blueshifted by $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$5; each component had characteristic widths of order a few $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$6, and the velocities remained stable over $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$7 weeks (Perley et al., 6 Jan 2026). Proposed interpretations included tidal debris streams, an ablated companion star, or other stable asymmetric structures. The late appearance of only H and He features indicates that the line-forming region differed physically from the highly ionized continuum-forming wind.

4. Polarimetry and constraints on ejecta geometry

AT2024wpp is only the second Cow-like transient with optical polarimetry. Multi-epoch BVRI imaging polarimetry obtained with NOT/ALFOSC between $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$8 and $$\mathrm{SFR}\approx0.075\pm0.010\,M_\odot\,\mathrm{yr^{-1}}$$9 d showed a degree of polarization $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$0 across all bands and epochs, consistent within errors with $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$1 (Pursiainen et al., 2024). Stokes $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$2–$\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$3 measurements clustered near the origin and showed no significant preferred position angle.

The low measured polarization was interpreted as intrinsic rather than an artifact of interstellar polarization. Along the Galactic line of sight, the reddening is $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$4, which via the Serkowski relation gives $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$5, or a Milky Way ISP of $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$6. Nearby stars show $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$7, and the transient showed no host Na I D absorption and a blue continuum, both supporting low host extinction and therefore low host ISP (Pursiainen et al., 2024). The argument advanced in the observational paper is that it is very unlikely that Galactic and host ISP would cancel a substantial intrinsic polarization so precisely.

For an electron-scattering photosphere, continuum polarization near $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$8 corresponds to only modest asphericity, with an illustrative oblate-spheroid axis ratio $\mathrm{SFR}\approx0.1\,M_\odot\,\mathrm{yr^{-1}}$9 (Pursiainen et al., 2024). Since AT2018cow likewise showed nearly zero intrinsic polarization during a comparable outflow-dominated phase, the combined evidence was taken to argue that the high-velocity outflows themselves are likely highly spherical, at least in the layers that host the optical photosphere.

This conclusion does not eliminate inner asymmetry. AT2018cow showed an earlier high, wavelength-dependent polarization phase, reaching $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$0, attributed to shock breakout through a thin equatorial disk, before rapidly dropping to near zero (Pursiainen et al., 2024). AT2024wpp was not observed polarimetrically before $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$1 d, so a similarly brief early asymmetric phase could have been missed. A stratified geometry was proposed in which a dense equatorial torus or disk persists near the engine while a fast, roughly spherical outflow of order $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$2 expands around it; in that picture, early asymmetry and later near-sphericity are not contradictory (Pursiainen et al., 2024).

5. Radio, X-ray, and minute-timescale variability

Reports of strong X-ray and radio emission were central to the classification of AT2024wpp as a Cow-like transient (Pursiainen et al., 2024). VLA measurements around 29 days post-discovery detected the source at 10 and 15 GHz with flux densities of $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$3 and $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$4, corresponding to a 10 GHz specific luminosity of $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$5, a few times higher than AT2018cow at similar epoch and frequency (Ofek et al., 25 Aug 2025). Swift-XRT and Chandra observations showed an early soft X-ray source with roughly constant flux for about a week, followed by decay; around 50 days after discovery, a rapid X-ray brightening over a few days and substantial spectral hardening were reported (Ofek et al., 25 Aug 2025). A broader campaign later incorporated 58 Swift/XRT observations over the first $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$6 days, as well as VLA and ALMA coverage from weeks to months after explosion (Perley et al., 6 Jan 2026).

Radio and millimetre modeling indicated a mildly relativistic external shock. Using synchrotron self-absorption fits, one analysis inferred $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$7 and $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$8 at 30 days, $12+\log_{10}[\mathrm{O/H}] \approx 8.4\text{–}8.6$9 and $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$0 at 40 days, $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$1 and $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$2 at 65 days, and $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$3 and $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$4 at 107 days (Perley et al., 6 Jan 2026). The inferred density profile declined approximately as $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$5, and the shock energy under equipartition was of order a few $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$6 (Perley et al., 6 Jan 2026).

AT2024wpp also became a test case for whether AT2022tsd-like minute-timescale optical flares are generic among Cow-like transients. A targeted LAST campaign monitored the source from 28 to 74 days after the approximate zero-flux time. The paper reports about 155 hours of visits in total, an effective non-overlapping monitoring time of $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$7 hours in 6-min resolution, and an analysis in 10-min bins sensitive to flares with $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$8 at $E_{B-V}<0.034\text{–}0.12\,\mathrm{mag}$9; the abstract characterizes the observations as about 23 hours at the relevant sensitivity (Ofek et al., 25 Aug 2025). No minute-scale optical flares were detected. For $\mu \simeq 38.06\,\mathrm{mag}$00-min flares at that luminosity threshold, the reported $\mu \simeq 38.06\,\mathrm{mag}$01 upper limits were $\mu \simeq 38.06\,\mathrm{mag}$02 and $\mu \simeq 38.06\,\mathrm{mag}$03 (Ofek et al., 25 Aug 2025). These limits are well below the $\mu \simeq 38.06\,\mathrm{mag}$04 duty cycle and $\mu \simeq 38.06\,\mathrm{mag}$05 flare rate observed in AT2022tsd, indicating that such flaring is not universal within the class.

The flare non-detection was interpreted in three broad ways: intrinsic diversity in the Cow-like population, viewing-angle or beaming effects, or continued optical depth $\mu \simeq 38.06\,\mathrm{mag}$06 along the line of sight during the search window (Ofek et al., 25 Aug 2025). A plausible implication is that short-timescale variability, unlike gross UV–optical and radio phenomenology, is sensitive to geometric and radiative-transfer details rather than being a class-defining property.

6. Physical interpretations, modeling tensions, and broader significance

AT2024wpp has been used to test several competing physical frameworks: engine-driven supernovae, interaction-powered supernovae, shock-cooling models, tidal disruption events by intermediate-mass black holes, and exotic merger channels including Wolf–Rayet/black-hole systems (Pursiainen et al., 2024, Omand et al., 6 Jan 2026). The observational papers emphasize that its hot continuum, rapidly receding photosphere, near-relativistic outflow signatures, and strong radio/X-rays are difficult to reconcile with a canonical supernova powered primarily by radioactive heating or simple diffusion through static ejecta (Pursiainen et al., 2024, Perley et al., 6 Jan 2026).

One detailed interpretation favors rapid accretion onto a black hole. In that picture, a powerful central engine drives a fast wind with $\mu \simeq 38.06\,\mathrm{mag}$07, irradiates the ejecta and circumstellar medium with X-rays, and produces a quasi-spherical, optically thick reprocessing layer whose radius first expands and then recedes as the mass-loss rate declines (Perley et al., 6 Jan 2026). The same study argues that the high Doppler velocities and intense ionization suppress both optical and far-UV line formation, and it treats the late double-peaked H and He lines as signatures of denser, more stable structures such as tidal streams or an ablated companion. If AT2024wpp is interpreted strictly in a TDE-like framework, that work gives a model-dependent upper limit $\mu \simeq 38.06\,\mathrm{mag}$08 (Perley et al., 6 Jan 2026).

A separate multiwavelength fitting analysis reached a more skeptical formal conclusion. Using semi-analytic models for the optical, radio, and X-ray light curves, it found that none of the tested models could reasonably explain all aspects of the data simultaneously (Omand et al., 6 Jan 2026). Supernova-like and shock-cooling models could mimic the luminosity evolution only by invoking very high photospheric floor temperatures; when forced to cool more realistically, they underpredicted the UV flux. The tested TDE prescriptions evolved too slowly or had rise shapes that did not match the observed $\mu \simeq 38.06\,\mathrm{mag}$09-day rise. Radio and X-ray fits also proved mutually inconsistent under single-component jet or blast-wave descriptions (Omand et al., 6 Jan 2026).

Despite that failure of simple global fits, the same modeling study argued that physical considerations still favor a stellar-mass/IMBH TDE of a low-mass star plus a synchrotron blast wave in dense circumstellar material (Omand et al., 6 Jan 2026). Using late-time UV and X-ray limits, it excluded large regions of the $\mu \simeq 38.06\,\mathrm{mag}$10–$\mu \simeq 38.06\,\mathrm{mag}$11 plane, including $\mu \simeq 38.06\,\mathrm{mag}$12 at high confidence for any disc mass, and $\mu \simeq 38.06\,\mathrm{mag}$13 for $\mu \simeq 38.06\,\mathrm{mag}$14 (Omand et al., 6 Jan 2026). The resulting picture is not a settled identification but rather a constrained landscape: a compact engine appears necessary, black-hole accretion is favored in at least one interpretation, but no current semi-analytic model reproduces the full broadband dataset without tension.

In population terms, AT2024wpp occupies the extreme luminous end of the LFBOT distribution. One study states that ZTF found three LFBOTs at $\mu \simeq 38.06\,\mathrm{mag}$15 in seven years, implying $\mu \simeq 38.06\,\mathrm{mag}$16, or $\mu \simeq 38.06\,\mathrm{mag}$17 of the core-collapse supernova rate, and places an upper limit of $\mu \simeq 38.06\,\mathrm{mag}$18 on events as luminous and fast as AT2024wpp (Perley et al., 6 Jan 2026). Its significance therefore extends beyond a single transient: it constrains how engine power, circumstellar structure, line suppression, and geometric symmetry can coexist in one of the rarest known classes of explosive transients.