Benchmarking work was recently performed for the issue of molten corium concrete interaction (MCCI). A synthesis is given here. It concerns first the 2D CCI-2 test with a homogeneous pool and a limestone concrete, which was used for a blind benchmark. Secondly, the COMET-L2 and COMET-L3 2D experiments in a stratified configuration were used as a post-test (L2) and a blind-test (L3) benchmark. More details are given here for the recent benchmark considering a matrix of four reactor cases, with both a homogeneous and a stratified configuration, and with both a limestone and a siliceous concrete. A short overview is given on the different models used in the codes, and the consistency between the benchmark actions on experiments and reactor situations is discussed. Finally, the major uncertainties concerning MCCI are also pointed out.