Floating macroalgae blooms, primarily caused by <em>Ulva prolifera</em> and <em>Sargassum horneri</em>, frequently occur in the marginal seas around China, posing significant ecological and economic challenges. However, a long-term, high-frequency monitoring dataset capable of capturing bloom seasonality, interannual variability, and rapid transport dynamics across multiple sea regions remains limited. In this study, we derived 1-km weekly mean <em>U. prolifera</em> and <em>S. horneri</em> dataset covering 2018 to 2025, generated from 85,868 multi-sensor satellite images acquired from the Moderate Resolution Imaging Spectroradiometer (MODIS), Visible Infrared Imaging Radiometer Suite (VIIRS), and Ocean and Land Colour Instrument (OLCI). By integrating complementary observations from multiple satellite sensors, the dataset substantially improves temporal continuity and observation availability relative to single-sensor products, enabling more complete characterization of rapidly evolving macroalgal blooms. A two-step deep learning-based approach was employed to detect and distinguish <em>U. prolifera</em> and <em>S. horneri</em>, achieving >90% accuracy for both bloom extraction and species classification across sensors. Independent validation using hyperspectral Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) Ocean Color Instrument (OCI) and Sentinel-2 MultiSpectral Instrument (MSI) data further supports the reliability of species discrimination. The pair-wise cross-sensor biomass comparisons show strong consistency, with <em>R²</em> > 0.8 for both species. The dataset reveals distinct spatiotemporal patterns of the two macroalgal species. <em>U. prolifera</em> blooms in the Yellow Sea exhibit a relatively consistent seasonal cycle from May to August, with rapid aggregation and dissipation and the highest biomass recorded in 2021, whereas <em>S. horneri</em> shows a broader and more variable distribution across the East China Sea, with record-high biomass observed in 2024. Weekly observations further resolve rapid bloom aggregation, transport, and dissipation processes that are difficult to capture using monthly composites. Weekly products provide a balance between temporal resolution and observational completeness, preserving short-term bloom variability while supporting consistent interannual comparisons. Compared with weekly products, monthly composites underestimated seasonal peak biomass by 24–50% for <em>S. horneri</em> and 12–52% for <em>U. prolifera</em> during 2018–2025. In contrast, daily products were more sensitive to short-term biomass fluctuations, which could alter the identification of the year with the highest boom intensity, highlighting the importance of temporal aggregation for robust interannual comparisons. Comparisons with previous studies and reported bloom events demonstrate that the dataset reliably reproduces observed spatial distributions and seasonal characteristics. By providing a high-resolution, weekly, and species-resolved dataset of floating macroalgae, this study offers a consistent observational foundation for investigating species-specific bloom dynamics, assessing their responses to environmental variability, and enabling data-driven management and early warning of macroalgal blooms in China’s marginal seas.