The authors present the biogeochemical cycle implemented in CLIMBER-X. This model is one of the few models of intermediate complexity to simulate Earth system changes over many thousands of years. The model includes modules to simulate marine biogeochemical tracers, ocean-sediment interactions, and weathering as well as representation of the land biosphere, methane, and wetlands.  This publication is timely and will serve the group and the community as a reference. The manuscript is generally clear and the authors manage to present a rich set of model outcomes. I recommend publications after the following comments have been addressed.

Representing ventilation time scale reasonably well is a prerequisite for simulating biogeochemical tracers in the ocean. Natural and bomb-produced radiocarbon, and CFCs are often used to probe the ventilation time scales of the ocean. The authors should present how well these different age tracers are simulated by the model.

a) Results for CFC-11 and CFC-12 are not shown and discussed though mentioned in the model evaluation section 4 on line 290. Please compare simulated versus measured CFC-11 and CFC-12 distributions and inventories in additional figures and text.

b) The marine radiocarbon distribution is presented at the very end of the ocean section (Fig. 20). However, age biases are key to understanding the biases in the various biogeochemical tracers. Please show and discuss radiocarbon results before the discussion of the other tracers.

c) Please present and discuss the distribution of bomb-produced and natural radiocarbon separately

d) The discussion of biases in biogeochemical tracers (DIC, ALK, P, O2, Si, d3C…) would benefit by linking these biases to the biases in the age tracers.

I miss a table that provides the model parameters. It would be very helpful if model parameters and key equations were summarized in a table (could also be in an appendix or as SI)

Iron limitation is prominent in the model (Fig.7). The authors should provide more detail on the iron cycle implemented in the model. How is the ratio of aeolian input versus advection of Fe in the Southern Ocean and elsewhere?  How do the different iron sources of the model ocean compare with each other and other estimates? What is the role of ligands? Which parameters have been applied? What fraction of deposited iron is bioavailable? How will the balance between aeolian versus marine sources affect the model’s sensitivity to glacial-interglacial dust deposition? Part of this information could be nicely incorporated in table 1.

The model applies a temperature-sensitive particle remineralization rate. However, viscosity and thus particle sinking are also influenced by temperature. These two factors have opposite impacts under changing temperatures.  Why is the temperature dependence of viscosity not considered? Will this cause a too-large sensitivity of atmospheric CO2 to changes in ocean temperature in glacial-interglacial simulations?

Conclusions: It would be nice if the authors present an outlook on future, planned (?) model improvements. For example, the implementations of flexible stoichiometry instead of constant Redfield ratios or N2O in the land biosphere appear to be possible targets.

Title: The model remains an Earth System Model of Intermediate Complexity and this should be reflected in the title. Please replace the term Earth system model with Earth System Model of Intermediate Complexity.

L115: Please be more specific about how this virtual flux approach is implemented. Is the global net surface flux set to zero? How is the dilution effect implemented during times of net global freshwater addition/removal during ice sheet melt and formation?

L123: Could you please explain how this integration works? I guess HAMOCC in the ESM has a time step of order 30 min and is resolving daily radiation changes. How does the scheme account for sub-daily fluctuations in radiation for computing photosynthesis?

L148: In CLM4.5 photosynthesis is downregulated by nitrogen limitation on a daily basis. This misconception/flawed approach has been corrected in CLM5. Why are you following this approach? It is strongly recommended to update to a more realistic N-limitation or is the model here running without N-limitation?

L174: Are there pools for products (paper, wood for construction) fed by deforestation? Please explain.

L223: d13C of fossil fuel has changed considerably over the industrial period. Assuming a constant value of -26 permil leads to a positive bias. Please prescribe d13C of fossil emissions using updated information, e.g., Andres et al.

L 293: Please comment on whether changes in C3 and C4 crops are prescribed for d13C discrimination.

Fig. 2c I am a bit surprised that the DIC inventory at time 0, at the beginning of the spin-up is somewhat lower than the best estimate from the GLODAP data (37400 PgC), despite prescribing the DIC distribution from GLODAP at the beginning of the spin-up. Is the ocean volume too small?

Fig. 13: Perhaps specify the boundary of the SO and whether the SO section is included in the profiles of the Atl., Pac., and Indian.

Figs. 14, 15, 17, 18,  19, 20:  Why is the Southern Ocean sector only shown for the Indian Ocean? It would be much more instructive to show the SO sector of the Atlantic and Pacific in the top rows.

How are weathering fluxes distributed in the ocean?