The effectiveness of biochar as a soil amendment is highly dependent on local physical and chemical soil properties. Although the literature has already addressed biochar in several studies, there are still knowledge gaps. One the one hand, the relevant studies have primarily focused on field trials and small-scale applications at regional levels, overlooking the global perspective and regional differences. One the other hand, geospatial assessments lack quantitative evaluations and explanation, which are crucial for the model's applicability and the optimisation of biochar supply chains. Thus, this study addresses this gap by examining the impact of biochar on agriculture at a global scale. First, correlations between climate, soil, and fertiliser data, and maize yield are derived through Random Forrest machine learning algorithm. Subsequently, the relevant soil properties are adjusted to simulate the potential changes upon implementing biochar. Finally, the model projects the estimated maize yield following the introduction of biochar. Our findings demonstrate diverse effects of biochar, with notable increase in maize yield in arid regions of Africa and Asia. A substantial increase in maize yield is particularly expected in regions with a high bulk density, as biochar effectively loosens the soil, and in areas with a low soil organic carbon content, which is enhanced by biochar. Contrariwise, in northern South America, Central and North America, South-East Asia, and parts of Europe show low potential or even maize yield decreases. The model was also validated by comparing the results with 8 field trials from different countries, demonstrating a high level of accuracy. The outcomes are crucial for optimising biomass utilisation pathways, as it predicts the impact of biochar in different regions. Consequently, policy frameworks can be tailored to encourage biochar use in agriculture, especially in regions with the highest potentials, to fully leverage its sustainability and productivity benefits.