Soil stabilization using materials of biological origin is gaining acceptance and is widely promoted to encourage sustainable practices in geotechnical engineering. Biopolymers (BP) of various types have been researched extensively in the past decade for use as eco-friendly and low-carbon footprint additives for soil stabilization and have shown appreciable improvement in treated soil. The strength of the BP-treated soil primarily depends on soil type, biopolymer dosage (BD) and curing period (CP). This study proposes a non-linear model for predicting the strength of clayey sand as a function of BD and CP. Xanthan gum is used to improve the strength of the soil, a sand-clay mixture. Undrained shear strength of BP treated at dosages of 0.25, 0.5, 0.75, 1, 1.25, 1.5, 1.75, 2, 3 and 5% were determined at curing periods of 7, 14, 28, 56 and 90 days experimentally. These results were used to construct a non-linear model to predict the strength of treated soil. The proposed model has a determinant accuracy (R2) of 0.98 indicating its appreciable performance. The model showed a minimum residual of − 16.82 kPa and 19.21 kPa with corresponding errors of − 12.4% and 9.09% respectively. Despite the fact that variation in strength remains within reasonable limits, there is also a need for a specific reliability analysis when BP is selected as a soil stabilizer.