We present upper limits on the axion-like particle (ALP) photon coupling constant $\gagg$ from a broadband spectral analysis of the Coma cluster (Abell~1656) radio halo. We have used flux density measurements spanning 30.9,MHz to 6.6,GHz, which probe ALP masses from $0.26,\muev$ to $55.9,\muev$, the widest continuous mass coverage from any single cluster to date. We model the synchrotron foreground with the physically motivated in-situ reacceleration spectrum, which provides a substantially better description of the strongly curved Coma halo spectrum than the simple power law adequate for the Bullet cluster. We adopt an aperture-matched profile-likelihood analysis as our primary statistical estimator, fitting the full 23-point radio spectrum simultaneously at each tested axion mass. For the more robust CMB-only and CMB plus extragalactic radio background (ERB) photon baths, our conservative baseline $2\sigma$ upper endpoints at $\ma=55.9,\muev$ are both $\simeq4.35\times10^{-8}\,\GeV^{-1}$. If the ICM kinetic temperature is additionally used as a Rayleigh–Jeans photon occupation following Ref.~, the same analysis gives $\gagg<7.33\times10^{-12}\,\GeV^{-1}$; we retain this much stronger value only as a conditional model benchmark and not as a robust constraint for the optically thin ICM. We further forecast the sensitivity achievable with the Square Kilometer Array (SKA) in its high-frequency bands. For an ideal 100,h SKA-MID Band~5b observation near band centre, the same thermal-noise calculation gives $\gagg\simeq4.5\times10^{-9}\,\GeV^{-1}$ for the CMB-only and CMB+ERB baselines, whereas the conditional ICM kinetic-temperature prescription gives $\gagg\sim7.5\times10^{-13}\,\GeV^{-1}$. A systematic comparison across 14 galaxy clusters demonstrates that the luminosity distance to the observer is the dominant figure of merit, with the nearby Ophiuchus cluster ($z=0.028$) emerging as the premier SKA-MID target in the same conditional forecast, at a projected limit of $\gagg\sim2.70\times10^{-13},\GeV^{-1}$.