The chopped basalt fiber reinforced concrete (BFRC) exhibits excellent tensile, flexural, crack resistance, and durability properties, but the factors influencing BFRC compressive strength have not been fully discussed. There are contradictory reports in some studies on the effect of basalt fiber (BF) on the concrete compressive strength. This study investigated the coupling influence of plain concrete and BF on BFRC compressive strength. Firstly, 328 data points from 22 literatures were collected, then a statistical analysis was conducted, indicating that when the water-binder ratio < 0.45, 0.4 < sand rate < 0.55, fly ash – binder ratio > 0.15, binder – aggregate ratio > 0.24, and recycled aggregate replacement ratio was low, adding appropriate BF tended to improve the compressive strength. To effectively evaluate BFRC compressive strength, XGBoost regression model and seven classification models were trained to predict the effect of BF on compressive strength. It is found that classification models have more practical values. Random Forest and XGBoost classification model exhibit well, with accuracy over 93%. Decision Tree performs conservatively, with the highest precision of “enhancement” and recall of “weakness” of 97.7% and 96.8%, respectively. The combined model of XGBoost and decision tree has 99% assurance rate. A smart design framework was developed, which could ensure that the compressive strength will not be weakened by the addition of BF, and provided adjustment strategies when the results were unsatisfactory.