Background: The lack of price transparency prevents healthcare purchasers from being able to negotiate lower pricing for their services. The negotiations that determine the costs patients pay is a confidential conversation between payers and medical providers. While the availability of data is limited, the Medicare Provider Utilization and Payment datasets provide insight into what providers are charging for their services. Methods: Models were created in Python (Decision Tree, Extremely Randomized Trees, and Random Forest) to develop a regression model that estimated the relative cost of submitted charges for hospital inpatient services. The model was developed using provider geographic location, ownership, national comparison of quality measures, and diagnosis-related groups to predict submitted covered charges. Results: It was determined that the Random Forest model performed the best with a MAE of $10,623.98, MSE of $629,712,972.24, RMSE of $25,094.08, and MAPE of 15.4%. The model also had a r-squared of 90.0%; that is, the 13 variables accounted for 90% of the variation seen on Average Submitted Covered Charges. The diagnosis-related group variables was determined to be the most important feature, followed by hospital location (City and State), Hospital Ownership, and Hospital Overall Rating. Conclusion: As healthcare costs continue to increase, transparency in price can result in decreased prices through increased competition and consumers having the knowledge to choose more affordable providers based on price comparisons . While barriers exist when it comes to transparency (lack of available data, patient behavior, etc.), the machine learning model and tool helps create the framework for additional improvements in this area of health services research.--Abstract.