Tuberculosis, caused by Mycobacterium tuberculosis, is a leading cause of infectious mortality. It affects humans and other mammals, including cattle, with significant health, ecological, and economic impacts. Understanding the drivers of its host-specificity shapes the landscape of infectivity, which could potentially inform strategies for tuberculosis management. We hypothesize that host-strain compatibility influences host responses to M. tuberculosis and we aim to reveal these transcriptional responses during in vitro infections. We infected human and bovine cell lines with two human-associated M. tuberculosis strains from lineages 5 and 6, as well as two animal-associated strains (Mycobacterium bovis and the Chimpanzee bacillus), and quantified infection ratios, cell death, and transcriptional responses. By integrating those data and comparing responses from different compatibility combinations, we identified global transcription profiles determined by strain-host compatibility. Our results show that the most compatible host-strain combinations had higher infection rates and a greater proportion of necrotic cells, indicating a more pathogenic phenotype in vitro. We also identified distinct transcriptional patterns for high and low compatibility infections. Both hosts had similar transcriptional responses to their corresponding strains, aiming to increase cell proliferation. This proliferation could favor bacterial growth, explaining the success of the strain in its corresponding host. Conversely, hosts responded to non-associated strains with a more effective immune response, among other strategies, explaining their lower success in the non-associated host. In conclusion, our hypothesis was confirmed, showing that bacteria-host compatibility determines common transcriptional responses, but also specific responses that depend on the infecting strain. This study enhances the understanding of host-specific adaptation mechanisms in different species.