CosmoFit: A Graphical User Interface for Bayesian Cosmological Parameter Estimation with User-Defined H(z) Models

We present CosmoFit, a desktop graphical interface for Bayesian parameter inference in background cosmology. The application is designed to simplify cosmological analyses by reducing the need to manually program each step, while maintaining the transparency and reproducibility of the results. Users can define an expansion history H(z), specify fixed and sampled parameters, assign prior ranges and proposal scales, select the available observational likelihoods, and run Markov chain Monte Carlo analyses from a unified graphical interface. CosmoFit uses Cobaya for Bayesian sampling and GetDist for posterior analysis and visualization. The current version supports cosmic-chronometer data and the official Pantheon+, Pantheon+SH0ES, and Union3 supernova likelihoods available through Cobaya. For each run, the application stores the configuration files, metadata, execution logs, chains, and numerical summaries required to inspect and reproduce the analysis. The performance of CosmoFit was evaluated through unit tests, a parameter-recovery test using synthetic data, and an independent comparison with an analysis performed directly in Cobaya. The posterior mean values obtained with both procedures agree to within less than one percent. In addition, the synthetic-data test recovers the fiducial parameters within the expected credible regions. As an illustrative example, we perform a spatially flat wCDM analysis using Pantheon+SH0ES and obtain marginalized constraints and triangle plots directly from the interface. Within its current scope, CosmoFit provides an accessible, transparent, and reproducible tool for research and academic applications involving explicit models of the background cosmological expansion.