The paper presents a NEMO-based model study of the Gulf of Finland (GoF) with two horizontal resolutions (0.5 km and 2 km). The authors discuss the impact of higher resolution on model performance and the role of convective parameterization in the thermocline. The paper is relevant to Ocean Science readers due to the importance of sub-mesoscale processes in the upper mixed layer and their influence on vertical transport in semi-enclosed seas like the Baltic. The study provides a tool for analyzing small-scale processes in the GoF. The paper is generally well-written, with a clear overview of model tests and results. However, major revisions are needed, particularly in the discussion of sub-mesoscale processes. The current discussion is too shallow, and the authors should address the limitations of data coverage by using longer simulations. The paper aims to provide insights into sub-mesoscale and basin-scale processes, but it lacks detailed analysis of sub-mesoscale features, such as small-scale eddies and spots with large vorticity, and their role in vertical transport. Specific comments include the need for a figure showing the model domain in relation to the Baltic Sea and a transect at 25.5E. The paper should compare its model setups with other Baltic Sea NEMO configurations, such as those by Hordoir et al. (2013, 2015). The reliability of boundary conditions from the HI-ROMB model at 23E should be addressed, as well as the versions of HIRLAM and HIROMB models used and their temporal resolution. The impact of lateral parameterizations on surface salinity and density gradients should also be discussed. The authors should clarify the validation of stratification using satellite SST data, as satellite measurements may not accurately represent the thermocline depth. The locations of east, west, and central profiles should be shown on separate figures. The model overestimates the turbocline depth, and this may also be spatially overestimated in higher-resolution simulations. The role of sub-mesoscale flows and their impact on offshore water exchange should be studied further. The authors should also reference previous numerical studies to support their claims of improved results. Minor corrections include fixing typos, replacing "paten" with "pattern," and ensuring consistency in figure legends. The authors should proofread the paper for proper English.The paper presents a NEMO-based model study of the Gulf of Finland (GoF) with two horizontal resolutions (0.5 km and 2 km). The authors discuss the impact of higher resolution on model performance and the role of convective parameterization in the thermocline. The paper is relevant to Ocean Science readers due to the importance of sub-mesoscale processes in the upper mixed layer and their influence on vertical transport in semi-enclosed seas like the Baltic. The study provides a tool for analyzing small-scale processes in the GoF. The paper is generally well-written, with a clear overview of model tests and results. However, major revisions are needed, particularly in the discussion of sub-mesoscale processes. The current discussion is too shallow, and the authors should address the limitations of data coverage by using longer simulations. The paper aims to provide insights into sub-mesoscale and basin-scale processes, but it lacks detailed analysis of sub-mesoscale features, such as small-scale eddies and spots with large vorticity, and their role in vertical transport. Specific comments include the need for a figure showing the model domain in relation to the Baltic Sea and a transect at 25.5E. The paper should compare its model setups with other Baltic Sea NEMO configurations, such as those by Hordoir et al. (2013, 2015). The reliability of boundary conditions from the HI-ROMB model at 23E should be addressed, as well as the versions of HIRLAM and HIROMB models used and their temporal resolution. The impact of lateral parameterizations on surface salinity and density gradients should also be discussed. The authors should clarify the validation of stratification using satellite SST data, as satellite measurements may not accurately represent the thermocline depth. The locations of east, west, and central profiles should be shown on separate figures. The model overestimates the turbocline depth, and this may also be spatially overestimated in higher-resolution simulations. The role of sub-mesoscale flows and their impact on offshore water exchange should be studied further. The authors should also reference previous numerical studies to support their claims of improved results. Minor corrections include fixing typos, replacing "paten" with "pattern," and ensuring consistency in figure legends. The authors should proofread the paper for proper English.