Thank-you for your thorough responses to the reviewer comments. Please proceed with submission of the revised manuscript. Your contribution to The Cryosphere is appreciated.
Chris Derksen

Dear Ann – the concerns of the reviewers are clearly addressed in the revised manuscript. The study provides a comprehensive assessment of CMIP6 model behaviour with respect to sea ice, and is a highly relevant contribution to The Cryosphere. I really like your additions to the manuscript, which now more clearly identify relationships between model formulation and ice budget terms. I have identified just a few minor remaining issues to address (note that line numbers refer to the clean version of the revised manuscript with no tracked changes). Thanks,
Chris Derksen

The level of detail with respect to model descriptions is inconsistent across the Appendix (e.g. very little information on CanESM). Can this be harmonized somewhat?
Line 124: consider changing to: “Snow ice: ice formation due to the transformation of snow to sea ice due to surface flooding”
Line 287: “If we consider the total amount of winter ice growth (here taken as the sum of the frazil and basal growth terms), the spread in modelled values is 3.9x103Gt, compared to the larger range of 5.9x103 Gt for the basal growth alone.” Can you provide a brief explanation for the apparent compensatory behavior for frazil formation versus basal ice growth?
Line 309: can you remove the phrase “likely to be”?
Line 341: “a simple scheme to account for the loss of drifting snow”. I think a reference should be added here to Lecomte, O., Fichefet, T., Flocco, D., Schröder, D., and Vancoppenolle, M.: Impacts of wind-blown snow redistribution on melt ponds in a coupled ocean – sea ice model, Ocean Model, 87, 67–80, https://doi.org/10.1016/j.ocemod.2014.12.003, 2014. 
Line 350: I don’t see a description anywhere of DFS versus CORE forcing?
Line 396: specify that “The remaining models have a minimum frazil ice thickness…”
Line 449: consider changing the header “Dynamics” to “Ice Advection” for consistency.
Line 556: perhaps add a statement that insight on ice advection from the projections is limited because analysis of winds and large-scale atmospheric circulation (and hence simulated ice motion) were out of scope for this study.
Line 591: “Overall, models with the largest decline in basal ice growth by the end of the 21st century tend to be those with the larger decline in winter ice cover.” In Section 6.1 and Figure 10c, you discuss and show the changes also as a function of ice unit area. Is there value in including this perspective also in Figure 12 (as noted on line 621)?