Nanopores are promising sensors for glycan analysis with the accurate identification of complex glycans laying the foundation for nanopore-based sequencing. However, their applicability toward continuous glycan sequencing has not yet been demonstrated. Here, we present a proof-of-concept of glycan sequencing by combining nanopore technology with glycosidase-hydrolyzing reactions. By continuously monitoring the changes in the characteristic current generated by the translocation of glycan hydrolysis products through a nanopore, the glycan sequence can be accurately identified based on the specificity of glycosidases. With machine learning, we improved the sequencing accuracy to over 98%, allowing for the reliable determination of consecutive building blocks and glycosidic linkages of glycan chains while reducing the need for operator expertise. This approach was validated on real glycan samples, with accuracy calibrated using hydrophilic interaction chromatography-high-performance liquid chromatography (HILIC-HPLC) and mass spectrometry (MS). We achieved the sequencing of ten consecutive units in natural glycan chains, which provided the first evidence for the feasibility of a nanopore-glycosidase-compatible system in glycan sequencing. Compared to traditional methods, this strategy enhances sequencing efficiency by over 5-fold. Additionally, we introduced the concept of 'inverse sequencing', which focuses on electrical signal changes rather than monosaccharide identification. This eliminates the reliance on glycan fingerprint libraries typically required in putative 'forward hydrolysis' strategies. When the challenges in both 'forward and inverse hydrolysis sequencing strategies' are addressed, this approach will pave the way for establishing a glycan sequencing technology at a single-molecule level.