Arboviruses, including the Zika virus, have recently emerged as one of the most important threats to human health. The use of metagenomics-based approaches has already proven valuable to aid surveillance of arboviral infections, and the ability to reconstruct complete viral genomes from metatranscriptomics data is key to the development of new control strategies for these diseases. Herein, we used RNA-based metatranscriptomics associated with Ion Torrent deep sequencing to allow for the high-quality reconstitution of an outbreak-related Zika virus (ZIKV) genome (10,739 nt), with extended 5′-UTR and 3′-UTR regions, using a newly-implemented bioinformatics approach. Besides allowing for the assembly of one of the largest complete ZIKV genomes to date, our strategy also yielded high-quality complete genomes of two arthropod-infecting viruses co-infecting C6/36 cell lines, namely: Alphamesonivirus 1 strain Salvador (20,194 nt) and Aedes albopictus totivirus-like (4618 nt); the latter likely represents a new viral species. Altogether, our results demonstrate that our bioinformatics approach associated with Ion Torrent sequencing allows for the high-quality reconstruction of known and unknown viral genomes, overcoming the main limitation of RNA deep sequencing for virus identification.