Sea ice is a key component of the Earth’s climate system as it modulates the energy exchanges and associated feedback processes at the air-sea interface in polar regions. These exchanges strongly depend on openings in the sea ice cover, which are associated with fine-scale sea ice deformations, but the importance of these processes remains poorly understood as most numerical models struggle to represent these deformations without using very costly horizontal resolutions ( 2 km). In this study, we present results from a 12 km resolution ocean–sea-ice coupled model, the first that uses a brittle rheology to represent the mechanical behaviour of sea ice. Using this rheology enables the reproduction of the observed characteristics and complexity of fine-scale sea ice deformations with little dependency on the mesh resolution. We evaluate and discuss the Arctic sea ice mass balance of this coupled model for the period 2000–2018. We first assess sea ice quantities relevant for climate (volume, extent and drift) and find that they are consistent with satellite observations. We evaluate components of the mass balance for which observations are available, i.e. sea ice volume export through Fram Strait and winter mass balance in the Arctic marginal seas for the period 2003–2018. The model performs well, particularly for the dynamic contribution to the winter mass balance. We discuss the relative contributions of dynamics and thermodynamics to the sea ice mass balance in the Arctic Basin for 2000–2018. Benefitting from the model's ability to reproduce fine-scale sea ice deformations, we estimate that the formation of sea ice in leads and polynyas contributes to 25 %–35 % of the total ice growth in pack ice from January to March, with a significant increase over 2000–2018. This coupled framework opens up new opportunities to understand and quantify the interplay between small-scale sea ice dynamics and ocean properties.