Open datasets of the world's largest freight transport hubs and operators
globalhubs.top publishes free, public-domain (CC0) datasets ranking the world's largest freight transport hubs — and the companies that move the cargo — across sea, air and land. Everything here is freight; passengers are out of scope.
What's inside
Six tidy, documented, versioned tables. Each is available as CSV and Apache Parquet, with a consistent schema (name, location, country/region, coordinates, the ranking metric, year-over-year change where known, operator, official website and postal address).
- Largest container ports — busiest maritime hubs by container throughput (TEU).
Hugging Face · Kaggle · GitHub · Zenodo - Largest cargo airports — busiest aviation hubs by air cargo (metric tonnes).
Hugging Face · Kaggle · GitHub · Zenodo - Largest inland freight hubs — dry ports and rail intermodal terminals by throughput (TEU).
Hugging Face · Kaggle · GitHub · Zenodo - Largest container shipping lines — ocean carriers by operated fleet capacity (TEU).
Hugging Face · Kaggle · GitHub · Zenodo - Largest air cargo carriers — airlines by cargo tonne-kilometres (CTK).
Hugging Face · Kaggle · GitHub · Zenodo - Largest freight railways — rail operators by net freight tonne-kilometres.
Hugging Face · Kaggle · GitHub · Zenodo
How it's built
Each ranking is generated from publicly available statistical sources by an open, repeatable pipeline, so anyone can reproduce it. Hubs are ranked by throughput; operators are ranked by volume — the leader-complete industry metric — rather than revenue: operated TEU for shipping lines, cargo tonne-kilometres for air cargo, and net tonne-kilometres for railways.
Figures are checked both automatically and by hand — public sources can be out of date, so
websites, operators and addresses are double-checked by a person. Undisclosed values are
marked n/d and never invented.
License & AI training
Everything is released under CC0 1.0 — dedicated to the public domain. Use it for anything, including AI / LLM training, with no attribution required (a link back to globalhubs.top is always welcome). Facts aren't copyrightable, and CC0 also waives the EU database right.
Permanence & citations
Every dataset is versioned and mirrored across Hugging Face, Kaggle, GitHub, Zenodo and the Internet Archive, and carries a permanent Zenodo concept DOI that always resolves to the latest version. The domain is held long-term, so links and citations keep working.
Corrections & additions
Spotted an error, or want a hub or operator added or a figure updated? Please open an issue on GitHub.
Author: Dmitry Golubnichiy (ORCID 0009-0007-3307-2202). Cite via the dataset DOIs above.