<p>We present pyDielectriX, an open-source Python package for modeling and analysis of frequency-dependent dielectric response data, particularly those obtained from broadband dielectric spectroscopy of materials. The package contains a comprehensive library of dielectric models—Debye, Cole–Cole, Cole–Davidson, Havriliak–Negami, and fractional cap–resistor frameworks—organized into modular Python classes. pyDielectriX offers curve-fitting routines based on nonlinear least-squares optimization using the Levenberg–Marquardt algorithm. Furthermore, the package implements Bayesian optimization to improve parameter initialization and mitigate convergence to local minima arising from poor initial parameter selection. To assist users in model discrimination, pyDielectriX incorporates a probabilistic selection strategy based on fuzzy Bayesian networks, enabling objective comparison among competing dielectric models. Additionally, the Bayesian information criterion is included as a complementary deterministic metric to evaluate model parsimony and penalize overparameterization. A graphical user interface is provided to enhance accessibility and adoption by the dielectric spectroscopy community. The package supports dual-domain analysis within the relative permittivity and electric modulus formalisms. Benchmarking using experimental datasets from different sources demonstrates that pyDielectriX enables accurate parameter estimation, improves reproducibility, and facilitates physically interpretable analysis of complex dielectric spectra. The source code and documentation are freely available on GitHub, promoting transparency, reproducibility, and community-driven development.</p>