This paper serves as the description,  justification and documentation of the simulations intended to be part of the AerChemMIP Phase 2 in CMIP7.  As such, it is a very descriptive paper and will indeed provide modeling centers the information they need to perform the simulations.  Most of my comments aim at clarifying the text in terms of its simulation details, but also in its justification of choices.   These should be addressed by the authors.

• Line 87: this is a very large data request. Is there a clear understanding and documentation that this is useful?  Maybe from AerChemMIP?
• Line 89: why just “idea”?
• Line 97: what is CACTI?
• Line 160: that is not a clear demarcation between chemistry packages. Even stratospheric chemistry focused packages have a representation of tropospheric chemistry (like methane oxidation and ozone production).  Complexity in HC chemistry is probably a better split
• Line 161: why “as much capability as possible”? Doesn’t that make comparison between model result more difficult? This should be justified
• Lines 170-172: this is very unclear. What is the point of using a version that has not done the DECK? It seems like you’re enabling using an unconstrained version that could be giving very different results (on say ECS) and not be documented
• Lines 212-214: Shouldn’t that be consistent with the way the model is going to be run for the historical simulations? Is the idea here that all changes in biomass-burning emissions are anthropogenic?  I don’t think that’s the case, especially in the boreal regions.
• Line 224: equilibrium in what?  This has to be better defined
• Line 264: if the “weak dependency” has already been documented, why should this setup be part of the proposed simulations?
• Section 3.1.3: it seems that this is not very well developed and specific requests should be made much clearer to be useful. In particular, documentation on the aerosol scheme (which is always a tricky one in the CMIP documentation) needs to be more complete than simple “interactive” or “prescribed”
• Line 311: don’t most models have some basic H2 chemistry? But most likely not enough to do it properly?  By the way are you asking for emission-based H2 simulations or surface concentration-based? 
• Line 333: Based on AerChemMIP, how many Tier 3 simulations were actually performed by enough models to make them useful to be part of CMIP?
• Line 350: What is the advantage of a SST+4K experiments over simply a future SST runs? I can see why it would be useful for CFMIP, not so much for AerChemMIP.  Justify why this should be in AerChemMIP2.
• Line 355: But, with that choice, the results will be a mix of models with and without biomass burning feedback? That will the intercomparison and attribution more complicated.  Why is that the better approach?
• Line 480: what is the rationale for a linear increase in dust emissions? That seems highly subjective and unclear why that would be a useful experiment
• Lines 488-490: Dust is actually a critical component to the PI-control tuning. How would this adjustment be consistent with that tuning?
• Lines 532-533: what is the relevance of that sentence to AerChemMIP2?
• Line 579: what is the meaning of “treatment of SLCFs”?
• Line 585: aren’t those all the SLCF emissions?
• Lines 603-604: is that the same linear extension described before? This is unclear
• Lines 734-735: why mentioned that those “could be explored”? If they are not, then they shouldn’t be in this paper.
• Lines 741-742: this is way too general to be useful. What diagnostics? Isn’t this what AeroCom does already?
• Line 762: even just one or two historical simulations is an enormous amount of data? Is that 3-hourly data? Who would run a CTM for the whole time-period?

This paper presents the AerChemMIP Phase 2 simulation design for CMIP7, clearly outlining its scientific goals and experimental framework. It is exciting to see the new experimental designs in AerChemMIP2 compared to Phase 1 and the scientific opportunities they enable. The paper provides detailed and practical guidance for modeling centers and offers a clear overview for future users of the resulting multi-model dataset. I only have a few comments:

1. Line 249, how is the nitrate aerosol treatment expected to change in CMIP7 models? Could the authors provide more details?
2. Line 277, the NMVOCs experiment is only designed for pdClim-X (table 2) but not for piClim-X (table1). The rationale for this choice is not clear and should be explained.
3. Line 345, the -2X experiments are designed for piClim but not for pdClim. Could the results be affected by state dependence if similar perturbations were applied under PD climatological conditions? Please consider adding a brief note discussing potential state-dependence issues here.
4. Line 392/398, “Aerosol precursors” and “O₃ precursors” are discussed as early in Table 1 but are not defined until this paragraph. I suggest defining these terms earlier in the paper. Also, it is somewhat confusing that NO_x and VOCs are categorized only as O₃ precursors but not as aerosol precursors, given that they also contribute to aerosol formation. Please clarify and justify this categorization. Relatedly, how are interactions between aerosol and O₃ treated or separated in the AerChemMIP2 experimental design?
5. Table 5, I suggest explicitly listing the reference experiments in the table (e.g., hist-piControl, esm-scen7-h, and esm-scen7-vllo). In addition, the reference for the −h and −vllo differs, but the same label (“Ref”) is used in the table, which may be misleading. Using distinct labels, as done in Table 8, would improve clarity. A similar issue also applies to Table 7.