Generally speaking, reaction events are probabilistic. Even in a magic world without equilibrium, and ignoring the possibility of side products, there's not a defined amount of time, t, after which every molecule in a sample is consumed to form products. Reaction kinetics are determined by rate laws and rate constants. From these values, which are largely determined empirically, you can predict the concentration of a reactant as a function of time. But this value typically approaches an equilibrium value asymptotically, meaning that if your criterion is for every molecule of a reactant to be consumed, the answer is that it will never happen in a precisely defined amount of time.
Photochemistry is a little different from the case described above because one of your reactants (the photons) is present in essentially infinite amount. This means that many photochemical reactions tend to be "zero order", and the concentration of reactant, barring something weird going on, is linear. Do note that photochemical reactions are still fundamentally probabilistic, but because the probability of a reaction event does not include the concentration-dependent probability of a collision, then the rate is more or less constant and the concentration of reactant decreases linearly. The slope of the line depends on such things as the molecular absorptivity and the efficiency of internal photophysical processes. Note that if the photon is only providing an activation energy but a collision with another reactant is still required, then the zero-order assumption may not be the case. Therefore you have to know a little bit about the reaction mechanism before you can make assumptions.
You may find this interesting:
https://pubs.acs.org/doi/pdf/10.1021/ed074p1303
ChemBuddy | 
Chem Reddit | 
Topic: Nitric oxide to Nitric dioxide conversion with UV (Read 2164 times)