Cloudflare renamed its Data Platform to Basin and made the Apache Iceberg analytics suite generally available on Pipelines, Catalog, and SQL

Cloudflare announced Basin today, the general-availability release of what it has been calling the Cloudflare Data Platform. The suite keeps the same three products and gets a new name. Basin Pipelines, formerly Cloudflare Pipelines, ingests events and writes them to R2. Basin Catalog, formerly R2 Data Catalog, manages Apache Iceberg metadata. Basin SQL, formerly R2 SQL, queries Iceberg tables directly on Cloudflare.

Cloudflare says it set out to build the platform a year ago, after it saw Apache Iceberg emerging as the standard open table format and developers moving analytics data to R2 for its lack of egress charges. The platform entered open beta during last year's Birthday Week and is generally available from today.

What the three products do

Basin Pipelines receives events from Workers, HTTP or Cloudflare Logpush, transforms them with SQL, and writes them as Iceberg tables or files in R2. Basin Catalog keeps the Iceberg metadata current and compacts data files on its own, so tables stay fast and cost-efficient as they grow. Basin SQL uses the catalog's statistics to plan a query, splits it into smaller tasks, and runs those tasks on Workers.

Cloudflare says a team can create a catalog, set up a pipeline, and run a query in seconds, and that this speed matters when a prompt or a coding agent would otherwise sit and poll for resources.

The Iceberg argument and the egress bill

Cloudflare makes the case that an open table format lets a team use the right query engine for each job: PyIceberg, DuckDB, Snowflake, Apache Spark or any other Iceberg-compatible engine can read the same data. The portability only works in practice because R2 charges no egress fees, so moving the compute to the data costs the same as moving the data to the compute.

Pricing is usage-based: a team is billed only when Basin ingests, processes or queries its data. Cloudflare does not quote a figure for a typical workload, and the three product pages are where the live numbers live.

Who is already using it

Cloudflare quotes two early adopters. Dax Raad, co-founder of Anomaly, says the company moved its entire data pipeline to Basin Pipelines, Catalog and SQL, replacing an AWS S3 and Athena setup. Julien Grobbelaar, head of platform at Bobsled, says the company uses Basin to build data products that stay accessible in any region of any major data and AI platform at a fraction of the cost thanks to zero egress fees. Both statements are the customers' own.

A team running analytics on another cloud today has an Iceberg-native alternative that charges only for ingest, compute and storage, with no fee for reading data out. The gap to measure is query speed on the team's own workload against whatever runs today, because the serverless split-and-distribute design has a different cost profile from a dedicated warehouse. Cloudflare's tutorial is the fastest path to a reproducible comparison; the vendor's own claims are a starting point, not evidence.

Source

• Primary source: Cloudflare: Introducing Cloudflare Basin, an open, serverless data platform, now generally available, 1 October 2026