<p>Noises drive heterogeneous neuronal firing by modulating firing activity via suppression or enhancement, enabling flexible encoding of input signals. Input or astrocytic current can also regulate graded firing heterogeneity, especially as astrocyte integrates neurotransmitters to fine-tune neuronal response precision. In this work, we investigate three types of noises in generating heterogeneous firing under the regulation of input current and astrocyte. Synaptic inputs approximated by Gaussian White Noise induce inverse stochastic resonance, characterized by initial noise-suppressed and subsequent enhanced firing rates. As input current increases, the suppression effect weakens and becomes confined to a narrower range of noise intensity. Gate and conductance noises, arising from stochastic opening and closing of ion channels, induce an exponential decay of firing rate with membrane area. Compared to the noiseless condition, both the firing rates with currentless-enhancement and current-suppression under gate noise, as well as the uniformly enhanced one under conductance noise, reduce disparity in firing rates induced by input current and diminish response sensitivity. Furthermore, across all three noise regimes, astrocyte–neuron coupling is biphasic: weak coupling increases firing rate whereas strong coupling suppresses it, with suppression linked to a spiking-to-bursting transition. These findings provide strong evidence for noise-induced heterogeneous firing and highlight the functional role of astrocyte and external inputs in neuronal information transmission.</p>