Skip to main content
The OpenAlex snapshot is hosted on Amazon S3 and is free to download. You don’t need an AWS account.
Many thanks to the AWS Open Data program, which covers the data-transfer fees (about $70 per download) so users don’t have to.

​
Prerequisites

Install the AWS CLI. All commands in this guide use it. No account or credentials are needed — the --no-sign-request flag provides anonymous access. You can also browse the snapshot in your browser: openalex.s3.amazonaws.com/browse.html

​
Download the full snapshot

This copies everything in the openalex S3 bucket to a local folder. It takes about 330 GB of disk space. 
aws s3 sync "s3://openalex" "openalex-snapshot" --no-sign-request
If you’re downloading into a folder that already has a previous snapshot, use the --delete flag to remove outdated files. Otherwise you’ll get duplicate entities that have moved between partitions.
aws s3 sync "s3://openalex" "openalex-snapshot" --no-sign-request --delete

​
Check the current size

The snapshot size changes over time. Check before downloading: 
aws s3 ls --summarize --human-readable --no-sign-request --recursive "s3://openalex/"

​
File structure

After downloading, you’ll have a structure like this: 
openalex-snapshot/
├── LICENSE.txt
├── RELEASE_NOTES.txt
└── data
    ├── authors
    │   ├── manifest
    │   └── updated_date=2024-01-15
    │       ├── 0000_part_00.gz
    │       └── 0001_part_00.gz
    ├── institutions
    │   ├── manifest
    │   └── updated_date=2024-01-15
    │       └── ...
    ├── sources
    │   ├── manifest
    │   └── ...
    ├── works
    │   ├── manifest
    │   └── ...
    ├── topics
    │   └── ...
    ├── publishers
    │   └── ...
    ├── funders
    │   └── ...
    └── merged_ids
        └── ...
See Snapshot data format for details on the partition structure and how to keep your copy up to date.

​
Download a single entity type

If you only need one entity type, specify the prefix: 
aws s3 sync "s3://openalex/data/works" "openalex-snapshot/data/works" --no-sign-request

​
Alternatives to local download

If you don’t want to download files locally, some services can read directly from S3:
  • Amazon Redshift: Load from S3 using the manifest files
  • ETL tools with S3 connectors (Xplenty, Airbyte, etc.)
For these approaches, the snapshot data format documentation should have enough detail to get started.