Background: Using long reads provide higher contiguity and better genome assemblies. However, producing such high quality sequences from raw reads requires to chain a growing set of tools, and determining the best workflow is a complex task. Results: To tackle this challenge, we developed CulebrONT, an open-source, scalable, modular and traceable Snakemake pipeline for assembling long reads data. CulebrONT enables to test on multiple samples multiple long reads assemblers in parallel, and can optionally perform, downstream assembly, circularization and polishing. It further provides a range of assembly quality metrics summarized in a final user-friendly report. Conclusions: CulebrONT leverages the difficulties of assembly pipelines development, and allow even basic users to obtain high-quality assemblies.