The development of neuromorphic computation has made enormous progress in time-series forecasting, non-structural visual data, and pattern recognition. Regarding hardware implementation, the temporal response of multiple inputs and processing-in-sensor techniques are especially crucial for the implemented sensing devices to manage massive data with nonlinear processing. In this study, we demonstrated an extensive conductance modulation of a lightsensitive zinc-tin oxide (ZTO) phototransistor in responding to an electrical− optical hybrid composed of a negative gate voltage pulse and 405 nm light pulse. By precisely controlling the lead time (Δt, −5 s ≤ Δt ≤ 5 s) of light spike in the hybrid pulse, a synergistic current amplification effect is revealed when the gate voltage spike and 405 nm light spike are fully or partially overlapped. When the phototransistor is embedded with gold nanoparticles (ZTO/Au NP phototransistor), the synergistic current amplification effect is even more significant. With Δt = 0 s, the fully overlain light/voltage hybrid pulse-induced current changes (ΔI h ) on ZTO and ZTO/Au NP phototransistors are 13-and 16-fold, respectively, of the value obtained by summing up the current induced by the individual pulse. The current amplification factor is spike-timing-dependent, and it is a multifaceted phenomenon correlating with the gate bias-enhanced electron detrapping and carrier generation in ZTO under light illumination, the hot electrons contributed from Au NPs, and the current retention that arise from the prior pulse. With the electrical−optical synergistic effect and manifold temporal response, both phototransistors show the potential to be implemented for data processing in sensor networks.