Power generation planning (PGP) is a vigorous stage in power system planning subsequently forecasting the load demand. The nature of the PGP problem is highly constrained, dynamic, discrete, nonlinear, stochastic, and mixed-integer. The amalgamation of the renewable energy sources in strategic PGP generates the complexity in terms of feeble reliability because of variability. The meta-heuristic algorithms are referred to as efficient solution techniques for the optimization of challenging and minimum-cost PGP. In this paper, a novel PGP optimizer has been proposed to plan the power generation plants reflecting the minimum cost and attaining an explicit reliability evaluation for specific planning prospect. The formulated and proposed optimizer is called teaching–learning-based optimization employing correction matrix method-with-indicators (TLBO-CMMI). In TLBO-CMMI, a new radix-5-based mapping procedure has been developed for the representation of population search agents. An innovative constraint handling procedure known as correction matrix method-with-indicators and incremental cost-based strategy to compute variable operation and maintenance cost, have been embedded from our previous research work. The optimizer has been employed to perform the PGP for different test scenarios in the literature including reliability and emission constraints. The optimizer gives promising results considering minimum cost and computational time when compared with results by current methods in the literature. The suitability of the optimizer has also been assessed by applying to an actual scenario of the Pakistan power system to create a viable power generation plan.