A fundamental challenge in the optogenetic toolbox is that all opsins, regardless of their excitation spectra, are activated by blue light. We hypothesized that pairing a red-shifted channelrhodopsin with a blue light-sensitive anion channel of appropriately matching kinetics shall render neurons responsive to a red but not blue light. To achieve this, we used a semi-rational mutagenesis strategy to optimize the kinetics and light spectrum of a chloride channelrhodopsin. By pairing optimized variants of blue light-sensitive anion channel ZipACR with vfChrimson, a fast red-shifted channelrhodopsin, we created a system in which red light derives precise and faithful action potentials of high frequencies, while blue light, through shunting inhibition, nullifies the effect of the red-shifted ChR2. Additionally, by a simple switch between red and blue lights, one can effectively excite or inhibit the activity of the same neurons.