Skip to main content
OpenAlex offers multiple ways to access data beyond the REST API. Choose the approach that best fits your use case.

​
When to use the API vs. downloads

Use the REST API

  • Quick lookups and searches
  • Building applications
  • Real-time data needs
  • Most use cases

Use data downloads

  • Large-scale analysis
  • Machine learning training
  • Building local search indexes
  • Offline access requirements

​
Download options

​
OpenAlex Snapshot

The complete OpenAlex database as gzip-compressed JSON Lines files, updated monthly. Includes works, authors, sources, institutions, topics, publishers, funders, and more. Best for: Full database replication, data warehousing, comprehensive analysis Size: ~330 GB compressed, ~1.6 TB decompressed Learn more about the snapshot format

​
OpenAlex CLI

The official command-line tool for downloading filtered subsets of OpenAlex data. 
pip install openalex-official
openalex download --api-key YOUR_KEY --filter "publication_year:2024" --output ./data
Best for: Downloading specific subsets without the full snapshot OpenAlex CLI documentation

​
Full-text PDFs

Download PDFs and TEI XML for about 60 million works. Requires an API key — content downloads cost $0.01 per file. Best for: Text mining, content analysis, building corpora Full-text PDF documentation

​
Decision tree

Do you need the complete database?
├── Yes → Download the snapshot
│         (/download/snapshot-format)
└── No
    ├── Do you need filtered metadata or content files?
    │   ├── Yes → Use the OpenAlex CLI
    │   │         (/download/openalex-cli)
    │   └── No → Use the REST API
    │             (/api-reference/introduction)
    └── Do you need bulk full-text PDFs?
        ├── Yes → See full-text PDF options
        │         (/download/full-text-pdfs)
        └── No → Use the REST API

​
Getting started

  1. For the snapshot: Follow the download instructions to get the data to your machine
  2. For the CLI: Install with pip install openalex-official and run openalex download --help
  3. For PDFs: See full-text PDFs for the three download options
Working with the full snapshot is challenging. The dataset is large (330 GB+ compressed) and complex. If you’re unsure, start with the REST API — it can answer most questions with much less setup.