Commercial vitamin-containing food products, such as food supplements, are increasingly consumed and manufactured through microbial fermentation. However, their biological purity remains largely unmonitored. Unexpected biological impurities with public health concerns were recently reported in similar fermentation-derived products, including pathogens, allergens and genetically modified microorganisms (GMMs) carrying antimicrobial resistance genes. Existing control strategies rely on targeted PCR-based methods, unsuitable to detect untargeted or unknown impurities. In this study, we developed an open shotgun metagenomic sequencing workflow for untargeted detection of biological impurities in vitamin-containing food products. This workflow integrates an optimized DNA extraction protocol tailored to diverse vitamin formulations and a novel bioinformatics pipeline, MetaCARP, compatible with short and long reads, enabling species-level identification and GMM detection. Workflow performance was evaluated using vitamin-containing food products artificially spiked with different biological impurities, including DNA from fungal, bacterial, plant and GMM sources. Results showed that high-level impurities were reliably detected, whereas detection of trace-level impurities remained limited. The applicability of the workflow was further assessed using two commercial vitamin B2 products previously reported as contaminated with unauthorized GMMs. The workflow not only confirmed the presence of known biological impurities but also uncovered unexpected ones, demonstrating the added value of a metagenomics-based open approach for impurity surveillance. This study provides a proof-of-concept for the use of shotgun metagenomics in food safety monitoring of microbial fermentation products, offering a comprehensive and flexible strategy that complements existing targeted control methods.