During additive manufacturing of stainless steels, sub-micron sized oxide (i.e., MnSiO3 , SiO2 , and CrMn2O4) and non-oxide (i.e., sulfide, in particular MnS, and possibly carbides, phosphides and nitrides) precipitates form during solidification. But do they evolve during the subsequent solid-state thermal cycling (SSTC) until the end of the printing process? A recent study on subjecting thin-film lamellae extracted from an additively manufactured stainless steel to heating-cooling treatments inside a transmission electron microscope (TEM) confirmed that precipitate composition can indeed evolve during SSTC. However, that study could not provide any conclusive evidence on precipitate volume fraction, density, and size evolution. In this work, we have quantified these changes using a novel experimental procedure combining (i) micropillar extraction from an additively manufactured stainless steel, (ii) subjecting them to different SSTC (including annealing) inside a TEM, (iii) performing synchrotron transmission X-ray microscopy to identify precipitates, and (iv) using a machine learning model to segment precipitates and quantify precipitate volume fraction, density, and size. Comparing these quantities before and after each SSTC/annealing sequence reveals that new oxides nucleated during rapid SSTC with high maximum temperature. However, during slow SSTC with high maximum temperature and annealing, precipitates dissolve because of oxygen evaporation during SSTC inside the TEM. A new empirical relationship correlating precipitate sizes and cooling rates is proposed. It is in good agreement with data collected from conventional casting, directed energy deposition, and powder bed fusion processes.