Supply chain finance, which solves the financing difficulties faced by small and medium-sized enterprises as a new model of financing, uses the credit granted by core enterprises to obtain financing from financial institutions. However, with the rapid development of supply chain finance, the linkage between financial markets is strengthened, as well as the number of financial entities and credit links increases, which leads to the credit risk of supply chain finance increasing. The traditional credit evaluation method of supply chain finance cannot solve the financing difficulties faced by small and medium-sized enterprises. To conquer these problems, an enhanced credit evaluation mechanism of supply chain finance based on blockchain technology is proposed in this article. Firstly the mechanism utilizes the characteristics of blockchain technology to solve the problem of information acquisition in the supply chain. Secondly, the random forest algorithm is used to perform feature processing on the relevant data, reducing the impact of redundant features on the model and improving the accuracy of the credit evaluation mechanism. Thirdly, combining the idea of stacking ensemble learning algorithm, a credit evaluation model is constructed (referred to the Huber-Stacking model). Finally, an opening dataset on a certain loan platform from small and medium-sized enterprise is used to verify the performance of the Huber-Stacking model. The experiment results show that the Huber-Stacking model has strong universality and can accurately evaluate company credit. The study shows that the supply chain finance credit evaluation mechanism based on blockchain technology proposed in this article can effectively resolve the financing difficulties faced by small and medium-sized enterprises and provide a reference for the construction of supply chain finance credit evaluation systems.