CMDFitter fits a probabilistic generative model to a color–magnitude diagram (CMD) of a star cluster or stellar population. Developed in Python, it uses a definition file to specify input photometry, isochrone models, and parameter priors, then samples the posterior distribution of stellar-population parameters (e.g., age, distance, metallicity) via nested sampling. The output includes posterior estimates of model parameters as well as diagnostic plots of the fitted model over-laid on the CMD. CMDFitter is particularly suited for interpreting deep stellar-population photometry in crowded fields and supports flexible customization of priors, isochrone sets, and observational uncertainties.