A catalog of the stellar atmosphere parameters of cool dwarfs in Gaia DR3, which is based on multi-band photometry from optical to infrared. The two files of .csv and .mrt contain the same contents. The information of CSV file is as follows: row number: 1,806,921. columns number: 24. columns names: column 1. 'source_id': (Gaia DR3 source_id). columns 2-3. 'RAdeg', 'DEdeg': (Right Ascension and Declination, decimal degrees). columns 4-11. ‘Mrmag', 'Mimag', 'Mzmag', 'MJmag', 'MHmag', 'MKsmag', 'MW1mag', 'MW2mag': (Absolute magnitude in multi bands). columns 12-17. 'Teff-RF', 'e_Teff-RF', '[M/H]-RF', 'e_[M/H]-RF', 'logg-RF', 'e_logg-RF': (parameters and errors by Random Forest algorithm). columns18-23. 'Teff-LGB', 'e_Teff-LGB', '[M/H]-LGB', 'e_[M/H]-LGB', 'logg-LGB', 'e_logg-LGB': (parameters and errors by LightGBM algorithm ). column 24. 'flag': (the boundary of BP - RP = 2.9 mag, flag = 1 while BP - RP > 2.9, otherwise flag = 0). The two tools in this paper are Random Forest (RF) and LightGBM, both of which are machine-learning algorithms based on decision trees.