Using stochastic simulation and chemical Langevin methods, we numerically investigated the influence of clustering of cells on the system size effects in two-dimensional coupled internal noisy cell systems. It was found that, when the system works near the Hopf bifurcation, the systems' cells assemble to a team in forming, a cell cluster. Interestingly, the teamwork in clustering may significantly raise the system's working efficiency, leading to the system obtaining the best performance under external stimulus. Furthermore, for different coupling intensities the optimal cluster size varies in robustness to external stimuli, but the signal-to-noise ratio (SNR) shows an increscent tendency with the coupling intensity increasing. These phenomena illustrate that clustering of cells may greatly enhance the sensitivity of Ca2+ signal, by which the biologic systems would probably improve their capability to respond to external stimulation.