Supermassive Stars Match the Spectral Signatures of JWST's Little Red Dots
The James Webb Space Telescope (JWST) has unveiled a population of enigmatic, compact sources at high redshift known as ``Little Red Dots'' (LRDs), whose physical nature remains a subject of intense debate. Concurrently, the rapid assembly of the first supermassive black holes (SMBHs) requires the formation of heavy seeds, for which supermassive stars (SMSs) are leading theoretical progenitors. In this work, we perform the first quantitative test of the hypothesis that LRDs are the direct observational manifestation of these primordial SMSs. We present a novel, first-principles pipeline generating synthetic spectra for a non-rotating, metal-free SMS up to $10^6 \, M_\odot$. We establish that its luminosity ($L_λ\approx 1.7 \times 10^{44} \, \text{erg} \, \text{s}^{-1} \, μ\text{m}^{-1}$ at 4050\,Å) provides a decisive constraint, matching prominent LRDs. Our model self-consistently reproduces their defining spectral features: the V-shaped Balmer break morphology is shown to be an intrinsic photospheric effect, while the complex line phenomenology, strong H$β$ in emission with other Balmer lines in absorption arises from non-LTE effects in a single stellar atmosphere. With wind and macroturbulent broadening, we match LRD spectra at $z=7.76$ and $z=3.55$, including the H$β$ width of MoM-BH*-1 to within 4\%. We predict a luminosity-dependent observability window, $\sim10^{4}$ yr for the most luminous systems and $10^{5}$--$10^{6}$ yr if $L_λ(4050\,\textÅ)$ is lower by 1--2 dex. These results provide a self-consistent alternative to multi-component obscured AGN scenarios and suggest JWST may be witnessing luminous stages of SMBH progenitors before collapse.