The Next-Generation Sequencing (NGS) platforms produce massive amounts of data to analyze various features in environmental samples. These data contain multiple duplicate reads which impact the analyzing process efficiency and accuracy. We describe Fast-HBR, a fast and memory-efficient duplicate reads removing tool without a reference genome using de-novo principles. It uses hash tables to represent reads in integer value to minimize memory usage for faster manipulation. Fast-HBR is faster and has less memory footprint when compared with the state of the art De-novo duplicate removing tools. Fast-HBR implemented in Python 3 is available at https://github.com/Sami-Altayyar/Fast-HBR.