TotalSegmentator review

Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026

TotalSegmentator is an open-source medical-imaging toolkit that automatically segments anatomical structures in CT and MRI data. Its whole-body CT task covers 117 classes, and its MRI task covers 50. Researchers and technical users can run it through a command-line interface, Python API, Docker container, or hosted web service. The hosted prototype is explicitly intended for research, not diagnosis or other medical use, so clinical users seeking a diagnostic tool should look elsewhere.

Its breadth of workflows makes it adaptable to research pipelines: the command-line interface offers selectable tasks and regions of interest, while the Python API accepts file paths or NIfTI objects. Docker supports container-based distribution, and the web workflow accepts anonymized CT or MR data. Outputs can include statistics, radiomics, previews, and DICOM segmentation. Training and evaluation guidance uses nnU-Net, making the project relevant to teams working with model development as well as pretrained segmentation. The web service limits uploads to one anonymized dataset up to 300 MB.

TotalSegmentator is free and open source, with cloud and self-hosted deployment options. Its published scope includes PyTorch and nnU-Net, and support is through documentation and the community. That combination suits research groups that can work with technical interfaces and manage their own imaging workflows. Commercial users should check licensing carefully: some model tasks require separate commercial-use licensing. Choose it for research-oriented segmentation and pipeline flexibility; avoid relying on the hosted prototype for medical decisions, or treating its open-source status as blanket commercial permission.

TotalSegmentator pros and cons

• Where it wins
  • Whole-body CT task segments 117 classes; MRI task covers 50 classes.
  • Offers command-line, Python, Docker, and web workflows.
  • Can produce statistics, radiomics, previews, and DICOM segmentation output.
• Where it doesn't
  • The hosted prototype is research-only, not for diagnosis or other medical use.
  • Web uploads are limited to one anonymized CT/MR dataset up to 300 MB.
  • Some model tasks require separate licensing for commercial use.

TotalSegmentator fact sheet, pricing and score →

Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. How we rank.

Last updated 9 Oct 2026 · How we research and update