metaSPAdes is a new versatile metagenomic assembler that addresses challenges in assembling complex microbial communities. Metagenomics is a key technology for analyzing bacterial populations, but assembly of metagenomic data remains difficult, especially with diverse and related bacterial strains. metaSPAdes leverages computational methods from single-cell and diploid genome assembly to improve metagenomic assembly. It was benchmarked against state-of-the-art assemblers and showed high-quality results across diverse datasets. Metagenomic assembly faces several challenges, including varying species abundance, interspecies repeats, and strain mixtures. These challenges are exacerbated by the high fragmentation of metagenomic assemblies, which affects binning accuracy and genome contiguity. metaSPAdes addresses these issues by focusing on reconstructing a consensus backbone of strain mixtures, ignoring rare strain features. The metaSPAdes pipeline constructs a de Bruijn graph from reads, transforms it into an assembly graph, and reconstructs long genomic fragments. It handles a wide range of coverage depths and balances accuracy and contiguity. It also uses a repeat resolution approach that utilizes rare strain variants to improve consensus assembly. metaSPAdes was benchmarked against IDBA-UD, Ray-Meta, and MEGAHIT on synthetic and real datasets. It outperformed other assemblers in scaffold length, gene prediction, and read alignment. It showed significant improvements in the SOIL dataset, which contains a highly diverse microbial community. metaSPAdes also performed well on the HMP and MARINE datasets. The software incorporates novel algorithms for efficient assembly graph processing, repeat resolution, and error correction. It is designed to handle large metagenomic datasets and incorporates emerging technologies like TSLR sequencing. metaSPAdes is available as part of the SPAdes toolkit and has been optimized for speed and memory usage. It is a valuable tool for metagenomic assembly, addressing key challenges in the field.metaSPAdes is a new versatile metagenomic assembler that addresses challenges in assembling complex microbial communities. Metagenomics is a key technology for analyzing bacterial populations, but assembly of metagenomic data remains difficult, especially with diverse and related bacterial strains. metaSPAdes leverages computational methods from single-cell and diploid genome assembly to improve metagenomic assembly. It was benchmarked against state-of-the-art assemblers and showed high-quality results across diverse datasets. Metagenomic assembly faces several challenges, including varying species abundance, interspecies repeats, and strain mixtures. These challenges are exacerbated by the high fragmentation of metagenomic assemblies, which affects binning accuracy and genome contiguity. metaSPAdes addresses these issues by focusing on reconstructing a consensus backbone of strain mixtures, ignoring rare strain features. The metaSPAdes pipeline constructs a de Bruijn graph from reads, transforms it into an assembly graph, and reconstructs long genomic fragments. It handles a wide range of coverage depths and balances accuracy and contiguity. It also uses a repeat resolution approach that utilizes rare strain variants to improve consensus assembly. metaSPAdes was benchmarked against IDBA-UD, Ray-Meta, and MEGAHIT on synthetic and real datasets. It outperformed other assemblers in scaffold length, gene prediction, and read alignment. It showed significant improvements in the SOIL dataset, which contains a highly diverse microbial community. metaSPAdes also performed well on the HMP and MARINE datasets. The software incorporates novel algorithms for efficient assembly graph processing, repeat resolution, and error correction. It is designed to handle large metagenomic datasets and incorporates emerging technologies like TSLR sequencing. metaSPAdes is available as part of the SPAdes toolkit and has been optimized for speed and memory usage. It is a valuable tool for metagenomic assembly, addressing key challenges in the field.