This manuscript presents and evaluates MOM6-COBALT-NEUS25 version 1.0, a high-resolution (1/25 deg.) coupled physical-biogeochemical model for the northeastern US designed to support marine carbon dioxide removal assessments and benthic modeling applications. The configuration consists of  the MOM6 ocean model coupled with COBALT using a z* coordinate with 75 vertical layers, incorporating tidal forcing, river discharge, and multiple data sources (Glorys, ERA5, TPXO9, WOA23) for initialization and boundary conditions. A thorough validation against observational databases, satellite products, and climatologies is presented for the simulation period in order to assess model performance. Key findings demonstrate that the model realistically simulates Gulf Stream dynamics, seasonal temperature and salinity structures with small SST biases, and realistic surface chlorophyll and nutrient seasonal cycles with expected spring and fall peaks. The configuration successfully captures regional circulation patterns and the development of cold SST patches through the inclusion of tidal mixing, an improvement over current reanalysis products. The work represents a significant advancement in regional ocean modeling by providing improved spatial resolution and process representation compared to global reanalyses, offering a robust foundation for climate impact studies and fisheries applications in this ecologically and economically important region. Some limitations include persistent biases in the Mid-Atlantic Bight's vertical temperature-salinity structure, a northward bias in Gulf Stream position with reduced meandering that affects tracer field representation, weak coastal freshwater signals leading to positive salinity biases, underestimation of mesozooplankton biomass, and limited skill in capturing subinertial sea surface height variability— however, many of these shortcomings arise from known modeling limitations and not necessarily from poor configurations choices. On the contrary, the authors have done a good job in with their parameter choices to ensure the best representation for their desired applications. The authors also present evidence on how these biases do not precluded the intended purpose of the configuration in its totality. This manuscript would be an excellent addition to the growing body of work on high-resolution regional modeling using MOM6. I would recommend this manuscript for publication after the authors consider the following comments:

1. The authors attribute the warm shallow water temperature biases during summer and fall partly to "overmixing associated with the upper ocean boundary layer scheme". Whereas Ross et al. (2023) (using ePBL alone) found that adjusting the submesoscale restratification front length parameter in the Fox-Kemper et al. (2011) scheme represented a key trade-off between mixed-layer depth and bottom-temperature bias in their configuration. Why was the hybrid approach selected over ePBL and the Fok-Kemper restratification scheme?. (1) what motivated the choice of the hybrid max(ePBL, KPP) scheme over ePBL with the Fox-Kemper restratification (as in Ross et al. 2023)? , and (2)  was the sensitivity of bottom temperatures to the restratification front length parameter identified in Ross et al. (2023) explored in NEUS25? (it is unclear from lines 6210-623 which paremeterizations were tested).

2. The time step specified in line 201 is inconsistent with the time step in Table 1. Please clarify.
3. Lines 603-605: The authors could compare with a higher resolution SST product like OSTIA which may improve known biases in OISST: https://data.marine.copernicus.eu/product/SST_GLO_SST_L4_REP_OBSERVATIONS_010_011/description
4. The colormap chosen for Figure 3 is inapropiate with the white background as the light blues are hard to distinguish from white.
5. I would suggest the authors to review when/where they are referencing figures. For example, the paragraph starting in line 366 does not reference any figure. I am aware that the paragraph is still discussing Figure 3, but I believe it is good practice to add an aditional reference in cases like this.
6. Several figures are lacking panel numbering which would make in-text referencing (which also needs revisions) much easier and effective for readers.
7. Figure 7 needs to be revised: top colorbar is unreadable, panels are not numbered and the shared axes don’t really work with the offset applied to each column.
8. Mooring/ADCP IDs in Figure 1 are different to Figure 4.

This paper describes a new simulation of an ocean model representing the North East US Continental shelf at 1/25° resolution, coupled to a biogeochemical model.  Most of the paper is devoted to documenting the fidelity of the simulated fields relative to various observational products.  These comparisons are fairly comprehensive, quantitative and clear; the figures representing the model-observation comparisons are also clear and attractive.  Therefore, the paper will serve as a valuable reference for future work that will then use the model to address scientific questions.  I support the publication of the paper after the authors address a few questions/suggestions below.

I just have two somewhat larger comments and then a few more minor points below that.

Major:

First, I think the authors should spend a little more time describing some basics of the model initialization and boundary conditions in Section 3.1.  Though I found this section helpful (and Table 1 usefully describes much of the model configuration), I found a few simple and fundamental details absent.  For instance, I can see that the model was initialized from Glorys in 1993 and used Glorys to provide boundary conditions. I suppose that means that the MOM6 simulation described here was also run from 1993-2019, but I can’t find this simple fact stated clearly.   Also, is the 6-month “spin-up” part of this period, but then excluded from the comparisons, or somehow external to this period?   More importantly and with respect to boundary conditions,  it is important to explain if there is any freshwater or salinity nudging to avoid drift. A couple of decades is plenty of time for the mean salinity (and even sea level) fields in the model to drift if the mass budget isn’t balanced (i.e. evaporation + precipitation + net freshwater flux across model boundaries).  Section 3 should explain whether/how the freshwater and overall mass budgets are balanced and, if not, whether that has led to domain-wide drift in salinity or sea level.  

Second, one rationalization given for developing this new simulation was to have a tool to evaluate mCDR processes, but no comparison of any simulated fields against the observed carbonate system was offered.  Given that there are GLODAP climatologies of TA, DIC and time-varying surface pCO2 fields derived from SOCAT data, I think this omission is strange and should be well-rationalized or remedied by including one or more carbon system comparisons.

Minor:  

Figure 3:  This depiction does not at all look like streamlines.  Rather, some of the lines seem to end abruptly (not possible for the non-divergent velocity field for which a stream function is defined).  Also, the “current speed” given in colors does not clearly correspond to the closeness of the streamlines, as would be expected.  I suspect this is not a classicly-defined stream function and should be clarified or replaced.

Figure 1 and Figure 4 - It’s not clear which stars correspond to which ADCPs. In other words is A01 = A? Why are there numbers in Figure 4 and not in Figure 1?

Line 385 - Here is where part of my “major point #1” above could be inserted.  Are salinities allowed to drift unrestrained?  Is the whole domain getting fresher? Saltier?

Figure 8 caption says “days” for the bottom right panel, while the colorbar is labeled “months.”  These should be consistent

Figure 11 caption and section 4.2.5:  I would guess that the data must be detrended before performing EOF analysis, otherwise the warming trend would dominate the PCs.  If they are detrended, that information should be added to the caption and/or text.

Figure 12:  I am curious (and suspect other readers would be too) why satellite SSH is not used as an observational field against which to compare simulated SSH.  It would be good to explain this choice somewhere.

Figure 14 discussion, around line 555 - I don’t find the explanation explaining the high chlorophyll in the Gulf Stream convincing and it would be good to either back it up with direct evidence (not necessarily a new figure, but at least a clear assessement of Gulf Stream nutrient concentrations and gradients at its edges).  At 0.25° resolution, WOA has  ~4 points across the 100 km Gulf Stream at its upstream boundary.  I’ve previously looked at the nutrient fields in WOA and their lateral gradients and recall that the fronts at the edges of the Gulf Stream seem to be  present.

Line 607.  I appreciate that throughout the paper, the authors mostly discuss comparisons quantitatively, eschewing descriptors.  However, here  “low bias” is stated with no quantification.  Adding a few quantifications in parentheses here would be appropriate.

Overall, the discussion and conclusion sections are brief, but this is a long paper with a relatively simple set of objectives, so that seems fine with me.  On the other hand, the discussion is almost exclusively focused on temperature, and, briefly, the relationship of temperature with chlorophyll and zooplankton.  The omission of any discussion here of salinity or sea level seems notable.  Adding 1-2 paragraphs to explain the fidelity (or differences) between these salinity and sea level changes  with observations would be appropriate.  If the paper comes to include carbon system comparisons, as I suggest would be beneficial, then those should be synthesized here too.  Finally, the conclusion reflects on the skill in the Gulf of Maine and new opportunities for interrogating the system using this simulations there.  It would also be useful to readers to understand limitations of the model and where attention could be focused to improve it.