The paper introduces QUAST-LG, an advanced tool for evaluating the quality of large-scale genome assemblies. It addresses the limitations of traditional evaluation methods, particularly in handling eukaryotic genomes with complex repeat structures and low coverage regions. QUAST-LG introduces the concept of an upper bound assembly, which estimates the theoretical limits of assembly correctness and completeness based on the given dataset and reference sequence. The tool also includes new quality metrics, such as k-mer-based completeness and correctness, which are more robust to the presence of transposable elements (TEs) and other specific features of eukaryotic genomes. The evaluation is conducted using six eukaryotic datasets from different sequencing technologies, including both short-read and long-read libraries. The results demonstrate that QUAST-LG provides more accurate and realistic assessments compared to conventional tools, and it can handle large genomes efficiently. The tool is freely available and can be used for both reference-based and reference-free assembly evaluations, making it a valuable resource for researchers in genomics and bioinformatics.The paper introduces QUAST-LG, an advanced tool for evaluating the quality of large-scale genome assemblies. It addresses the limitations of traditional evaluation methods, particularly in handling eukaryotic genomes with complex repeat structures and low coverage regions. QUAST-LG introduces the concept of an upper bound assembly, which estimates the theoretical limits of assembly correctness and completeness based on the given dataset and reference sequence. The tool also includes new quality metrics, such as k-mer-based completeness and correctness, which are more robust to the presence of transposable elements (TEs) and other specific features of eukaryotic genomes. The evaluation is conducted using six eukaryotic datasets from different sequencing technologies, including both short-read and long-read libraries. The results demonstrate that QUAST-LG provides more accurate and realistic assessments compared to conventional tools, and it can handle large genomes efficiently. The tool is freely available and can be used for both reference-based and reference-free assembly evaluations, making it a valuable resource for researchers in genomics and bioinformatics.