Run Spiral wherever you have bytes and compute.
Keep your tables in your own bucket, in the cloud and region you choose, or in Spiral-managed storage. Regardless of the storage provider, Spiral stores your data in open-source Vortex files. On Enterprise, BYOC runs Spiral’s managed compute in your cloud account, next to the bucket.
Spiral created Vortex and gave it to the Linux Foundation.
Spiral works with all your clouds.
On Scale, Spiral is a multi-tenant hosted service, and most of our customers already use it across more than one cloud. Enterprise adds single-tenant cells, BYOC in your own cloud account, or the topology that you choose. Tables live in Amazon S3, Google Cloud Storage, Azure Data Lake Storage Gen2, Cloudflare R2, Tigris, or Ceph, or in Spiral-managed storage. Spiral writes only under the root that you give it. It lists other files in place and never deletes them.
Runs on any cloud and object store
Your bucket or Spiral-managed storage
In your bucket, Spiral charges no storage fee. You can also keep tables in Spiral-managed storage, at $0.023 per GB-month.
Any cloud or region
Jobs read straight from the bucket, on any VM or container, in any cloud or on premises.
In your account
Your jobs read and write the bucket from their own processes on every plan. With BYOC on Enterprise, Spiral’s compaction and server-side queries run in your cloud account too.
Your identity
People sign in through your SSO. Pipelines use the workload identity that their platform issues.
Spiral holds metadata, not your rows.
When your tables are in your bucket, Spiral holds only names, versions, commit history, and grants. Its signing service sees only the method and path of each request, never the bytes. Each storage credential that it signs covers one table and lasts at most 15 minutes. A revoked grant cuts off storage access within those 15 minutes. With BYOC on Enterprise, Spiral can run this metadata in your own account as well.
Your tables stay open, and your keys stay short-lived.
Yes: Open Vortex files
DuckDB, DataFusion, and Spark read the files under your tables directly.
No: A proprietary file format
A proprietary format turns leaving into a migration project. Spiral’s files stay readable without Spiral.
Yes: An exit path
You can export any table to Iceberg.
No: Long-lived keys in your jobs
Bucket keys in job configs end up in logs and repos. Where a provider supports OpenID Connect federation, Spiral uses it and stores no key. Elsewhere, Spiral vends short-lived credentials, so your jobs never hold a bucket key.