With the increasing challenges faced by China's food logistics in multimodal transportation by road, rail, and water, information sharing and control have become key issues. A graded multi-chain traceability model for food multimodal transport based on blockchain is proposed to address problems with traditional food traceability systems, such as data centralization, lack of information transparency, and data forgery. This model includes a primary chain using Ms—Raft algorithm, a secondary chain using R—PBFT algorithm, and smart contract algorithms. The primary and secondary chains are anchored to each other through hash locking to ensure the reliability of data transmission. Comparative experiments show that the multi-chain model achieves a write throughput of 170 tps and read throughput of 280 tps, with average delays of 0.25 s for accounting and 0.50 s for writing, and an average query time of 0.41 s. Compared with the single-chain model, it offers improved network performance, consensus efficiency, and operational efficiency, with enhanced security, meeting the traceability requirements for food multimodal transportation.