Quantitative spectroscopy of single and multiple OB-type stars

Aims . The plethora of spectra of OB-type stars in observatory archives and the much larger numbers to come from the WEAVE and 4MOST spectroscopic facilities require efficient, but also accurate and precise methods for (semi)automatic quantitative analyses. Methods . Neural networks were used to emulate the spectra of single- and multi-star systems, trained on hybrid non-local thermodynamic equilibrium models that cover a wide range of atmospheric parameters and chemical compositions. To derive the full set of stellar atmospheric parameters and uncertainties, a Markov chain Monte Carlo algorithm was implemented to fit high-resolution spectra within 3000–10 500 Å. Results . The neural networks and fitting algorithm were bundled into a programme called Spectral Analysis Tool Using Restricted Neural networks (S ATURN ). In its current implementation, S ATURN facilitates the emulation of synthetic spectra for spectral types O7 to B9, which differ only negligibly from computed models. S ATURN was tested on a number of benchmark stars that have been studied before, including single OB stars and a detached eclipsing binary (DEB) system. Excellent agreement of atmospheric parameters and elemental abundances for up to ten metal species is found with respect to the data in the literature, often with reduced uncertainties of about 1% in effective temperature and 0.05 dex in surface gravity (1 σ values). For DEB components, the uncertainties are larger, in particular for the fainter secondaries when only a single-epoch spectrum is considered, and lower if surface gravities and the effective temperature ratio from previous detailed light curve and orbital dynamics studies are used as additional fit constraints. Uncertainties of elemental abundances are typically