Bulk data
Beyond the targeted access of the API and the website, OpenAlex offers a family of bulk data products: big buckets of data you download wholesale rather than query record by record.
The bulk-data family
| Product | What it is | Format | Freshness | Cost |
|---|---|---|---|---|
| Snapshot | The complete OpenAlex database — every entity | gzipped JSON Lines + Apache Parquet | Public releases roughly monthly; daily-refreshed copies on paid plans | Free (public S3 bucket) |
| Data feed & changefiles | Daily files of every record created or changed, to keep a local copy current | JSONL + Parquet | Daily | Paid plans |
| Content archive | Full-text PDFs (~60M) and Grobid TEI XML (~43M) for works | PDF, TEI XML | Continuously updated | Per-file via API/CLI; bucket sync on enterprise plans |
| OpenAlex CLI | Command-line tool for downloading filtered subsets (metadata and content) | JSONL | Live (pulls from the API) | Free metadata; content per-file |
Bulk or targeted?
- Use the REST API for lookups, searches, applications, and most analyses — it answers most questions with far less setup.
- Use the CLI when you want a subset on disk — “all works on this topic,” “these 50,000 DOIs” — without operating snapshot infrastructure.
- Use the snapshot (plus the data feed to stay current) when you need the whole database: data warehousing, machine-learning training, building your own search index, or offline analysis.
- Use the content archive for text mining and corpus building at full-text scale.
Warning: Working with the full snapshot is a real engineering project. The JSON Lines copy alone is ~750 GB compressed, and several terabytes decompressed. If you’re unsure, start with the API or CLI.
Licensing
OpenAlex metadata — the snapshot, changefiles, and everything the API returns — is CC0: public domain, free to use for any purpose. The full-text files in the content archive are different: PDFs retain their original copyright, and OpenAlex grants no additional rights to them.
Getting started
- Snapshot: see the snapshot reference, then follow the download recipe.
- Data feed: see Data feed & changefiles, then the changefiles recipe.
- Content archive: see Content archive for the three download paths.
- CLI:
pip install openalex-official, thenopenalex download --help(CLI reference).