A combination of Bayesian inference, physics modeling, and Markov chain Monte Carlo sampling allows for accurate inference of biomolecule numbers and their photophysical state in cellular clusters.