Gaia Data Release 3. Summary of the variability processing and analysis

Context. Gaia has been in operations since 2014. The third data release expands from the early (EDR3) 2020 by providing 34 months of multi-epoch observations that allowed us to probe, characterise and classify systematically celestial variable phenomena. Aims. We present a summary variability processing analysis photometric spectroscopic time series 1.8 billion sources done for DR3. Methods. used statistical Machine Learning methods sources. Training sets were built global revision major published star catalogues. For subset classes, specific detailed studies conducted confirm their class membership derive parameters are adapted peculiarity considered class. Results. In total, 10.5 million objects identified as DR3 have associated G, GBP, GRP and, some cases, radial velocity series. subdivide into 9.5 stars 1 Active Galactic Nuclei/Quasars. addition, supervised classification 2.5 galaxies thanks spurious induced extent these objects. output archive amounts 17 tables containing total 365 parameters. publish 35 types sub-types 11 types, additional object published. An overview estimated completeness contamination most classes is provided. Conclusions. Thanks we largest whole-sky based on coherent photometric, astrometric, data. Later releases will more than double span number observations, thus allowing an even richer catalogue future.