Pricing
The Rerun SDK is open source forever. Add Hub when you outgrow files on disk and need the data infrastructure run for you.
Hub replaces the export pipelines, custom dataloaders, and idle-GPU time you pay for today.
Open source
The full open-source toolchain on your own machine: log, query, transform, visualize, train, and add layers, all on one data model and a local catalog. Free forever.
Free
Apache-2.0 / MIT
Commercial
Everything in the SDK, run for you at scale in the cloud. The same APIs and workflows, now backed by a managed catalog over your own object storage.
Contact us
Single-tenant, in your region
Yes. The Rerun SDK is open source under Apache-2.0 / MIT. Use it in research and production, on the laptop and in the cluster, with no asterisks.
Hub is priced per-deployment based on team size and data scale. We work with each team on a contract that fits their footprint. Book a meeting and we'll walk you through it.
Rerun runs Hub for you, single-tenant, in the cloud region you choose. Your data stays in your own S3-compatible object storage and streams directly to your compute. For large teams with specific requirements we can deploy the data plane into your own account, but most teams do not need it.
Yes. We work closely with academic labs and physical AI research groups. Reach out and tell us about your project.
The SDK is the open-source toolchain you build your data loop with, free forever: log, query, transform, visualize, and train on one data model. Hub is the commercial backend that runs that same data layer at scale: catalog, byte-range indexing, schema management, and streaming over your own S3-compatible storage, run for you and single-tenant. The APIs are the same either way; Hub changes what runs underneath, not what you do.