Talk with our engineers by joining the new Onehouse community Slack!
Commercial
Pricing model
Onehouse runs a bring-your-own-cloud model and bills for compute consumed in OCU plus support, or a flat fee for large deployments. There is no volume fee and no storage fee. You pay your cloud provider directly for infrastructure, at your own negotiated rates, and the data stays in your account. The AI Gateway adds no charge on token volume either: you pay your inference providers directly, and Onehouse bills only for the compute the gateway runs on.
Consumption pricing. By default you pay for what you use. Consumption is metered from the instances running in your account, with several clusters bin-packed onto shared instances so you pay for fewer of them (scaling and resource sharing). You control spend through per-cluster OCU ceilings, a project-wide limit, sync frequency, and spot capacity for workers.
Flat-fee pricing. For large deployments, Onehouse also offers a flat annual fee for the full platform, Quanton included. The fee is set by your support requirements and the scale of the deployment. It often comes to a tenth of the total cost of ownership of Snowflake or Databricks for the same workloads, because a flat fee does not grow with every query or job the way their consumption pricing does. Contact Onehouse to discuss it.
Quanton on its own. For smaller deployments that want costs fixed and tied to data volume, Quanton is sold unbundled as a Spark engine, priced per GiB of data read and written, with the first 100 GiB each month free. See quanton.dev/pricing.
To model the savings before you commit, use the cost analysis tool or the Spark Cost Analyzer. Onehouse is also available through the AWS Marketplace.