Keeping global temperature rise well below 2°C, and pursuing efforts to limit it to 1.5°C, requires monitoring the greenhouse gases balance in the atmosphere. Carbon dioxide (CO2) contributes the most to global warming, and is emitted by fossil fuel combustion and land-use change, and taken up by the global land biosphere and the ocean. The Global Carbon Project (GCP) annually reports on global CO2 sources and sinks using process-based models and ocean data products. Furthermore, atmospheric inversions provide biosphere and ocean CO2 flux estimates based on atmospheric CO2 observations. In this paper, we present results from an ensemble of 14 inverse systems on both global and regional scales. We show that all atmospheric inversions close the carbon budget well on a global scale, but regional differences exist and the relative contributions of the land and ocean sinks to the total CO2 uptake is uncertain. We explore which regions contribute the most to these uncertainties. We also present metrics to assess the strengths and weaknesses of the inverse systems. Finally, we provide guidelines for using atmospheric inversions in future studies.