Pricing

One number to remember: how much you store.

Managed Apache Iceberg catalogs and storage. Your agents bring the compute; in a chat, we run the query.

Hosted in Frankfurt, Germany.

Free

For your own private data.

5 GB

No credit card required.

Start free

Public data Live

For open datasets anyone can query and join with their own.

Up to 50 GB

Free storage for the catalogs you make public: 5 GB as soon as you publish; up to 50 GB for datasets we review, on top of your 5 GB.

Publish a catalog

Pay as you go Coming soon

For private storage above the free 5 GB.

€0.10 per GB-month

The rate from the day paid storage opens; nothing else is metered. Charged monthly by card through Mollie (European), within a monthly budget you set; or by invoice and bank transfer for purchase orders.

Need more than 5 GB before then? Write to us and we raise your limit, free, until paid storage opens.

Ask for more storage

Enterprise

For organisations that sign in through their own directory.

Single sign-on with your OpenID Connect provider, SCIM provisioning, grants down to one table, service accounts, the audit log, managed in the EU or self-hosted, and support with a first answer within one business day. Priced in the agreement.

Talk to us What’s included

For your data, storage is the only thing we meter.

What you don’t pay for today.

  • No per-request fees

    Queries, commits and API calls.

  • No ingress fees

    Upload your data at no extra cost.

  • No egress fees

    Read your data back out at no extra cost, under fair use.

  • No compute charges

    Your engine runs the queries. Compute stays off our bill.

  • Maintenance included

    Compaction, snapshot expiry and orphan-file cleanup, every 6 hours.

  • No table or namespace fees

    No separate charge for each table or namespace.

Hosted agents (preview) are enabled per organisation on request during the preview, and are free while it lasts, within the plan’s run and token limits (4 runs a day and 5 million tokens a month per organisation); their pricing will be announced before the preview ends. Hosted agents →

Scheduled SQL is on for every organisation on the free plan, at no charge: up to two scheduled-SQL sinks per organisation (each keeps one table filled; a paused sink counts, a deleted one does not), run as often as every minute, each run within 60 seconds, 1,000 requests and 1 GiB downloaded. Need more sinks? Write to us. Scheduled SQL →

Your own compute provider may charge for the resources you use.

What you can do with LakehouseBox

  • A lakehouse for your company’s analytics

    Load your Parquet and CSV exports and query them with SQL, from your laptop or your agent.

  • Sensor and device data

    Devices and webhooks push rows straight into tables. No queue or server to run.

  • Photos and files from cameras

    Cameras and scanners drop files over FTP, FTPS or SFTP, straight into your catalog.

  • Memory your agents share

    Tables that persist between sessions and across agents: written with the CLI or your own engine, read through MCP too.

  • Maps and geospatial data

    GeoParquet in, geometry columns out, queried with spatial SQL, kept in spatial order by automatic spatial clustering.

  • Join your data with open data

    Query public datasets alongside your own tables in one query.

Works with DuckDB, PyIceberg, Spark, Trino and Polars · Iceberg REST catalog and S3-compatible storage · CLI, REST API, MCP and FTP · How it works →

Limits and fair use →Compared with S3 Tables, Snowflake, Cloudflare, MotherDuck →Read the documentation →

Start with your data. Bring your own compute.

Start free Questions? Contact us.