Stellar clusters are regularly used to study the evolution of their host galaxy .
Except for a few nearby galaxies , these studies rely on the interpretation of integrated cluster properties , especially integrated photometry observed using multiple filters ( i.e .
the Spectral Energy Distribution SED ) .
To allow interpretation of such observations , we present a large set of GALEV cluster models using the realistic approach of adopting stochastically-sampled stellar IMFs .
We provide models for a wide range of cluster masses ( 10 ^ { 3 } - 2 \times 10 ^ { 5 } M _ { \odot } ) , metallicities ( - 2.3 \leq [ Fe/H ] \leq + 0.18 dex ) , foreground extinction , and 184 regularly used filters .
We analyze various sets of stochastic cluster SEDs by fitting them with non-stochastic models , which is the procedure commonly used in this field .
We identify caveats and quantify the fitting uncertainties associated with this standard procedure .
We show that this can yield highly unreliable fitting results , especially for low-mass clusters .