Accurate modeling of bacterial biofilm growth is crucial for comprehending their complex dynamics in biomedical, environmental and industrial contexts. These dynamics are influenced by numerous environmental factors, such as antibiotics, nutrient availability and inter-species interactions. However, accurately representing these factors in computational models remains a challenge. In this contribution, we present an comprehensive multi-species continuum-based biofilm model designed to simulate a wide range of biofilm interactions with an arbitrary number of species. The biofilm model is derived from first principles of thermodynamics, ensuring a physically consistent framework. The model inherently satisfies the first and second laws of thermodynamics by the usage of Hamiltons principle of stationary action. The biofilm model captures the complex interactions among different species, nutrients, and antibiotics, and thus allow to simulate diverse growth behaviors, ranging from nutrient-deficient environments to dense multi-species communities. In addition, a Bayesian model updating framework is utilized to calibrate the computational model under hybrid uncertainties. The uncertainties are a combination of epistemic uncertainty (incomplete knowledge of the model parameters) and aleatory uncertainty (inherent biological variability and random environmental conditions). We demonstrate that not only the underlying model parameters can be robustly recovered even under sparse and noisy measurments but also predictive responses are provided.