Cloudflare takes its data platform out of beta as Basin

Cloudflare's data platform has left open beta under a new name. In a post published on 1 October, the company said Cloudflare Basin, the stack it launched a year ago as the Cloudflare Data Platform, is now generally available.

Three products, one pipeline

Basin is built on Apache Iceberg, the open table format, with data stored in R2 object storage, and it is made up of three products. Pipelines accepts events over HTTP endpoints or Workers bindings, transforms them with a SQL query and writes them to Iceberg tables; Cloudflare says ingestion now scales to 3GB/s per stream, up from the beta. Catalog is a managed Iceberg REST catalog that can be created with a single wrangler command and handles routine table maintenance, including compaction, as datasets grow. SQL is a serverless, distributed query engine for those tables, with hundreds of functions covering joins, aggregates and window functions, and no clusters to provision.

Open formats and no egress bill

The pitch leans on openness and cost. Cloudflare argues that Iceberg separates the storage layer from the compute layer so data stays portable between engines, and that pairing it with R2 removes the egress fees that usually sit on top of analytical workloads. Pricing is usage based, so customers are billed when Basin ingests, processes or queries their data rather than for capacity they are not using.

There is a customer quote attached. Julien Grobbelaar, head of platform at Bobsled, said Basin lets his team build data products that can be made accessible in any region, which is the sort of line a data platform launch lives or dies on.

Our opinion

The rename is the boring part. The interesting claim is that an edge network can host analytics next to where the data already sits, and Cloudflare is not pretending to out-feature the warehouse vendors: Basin's beta numbers look modest against a decade of dedicated tooling. What it does offer is a plausible answer to the egress bill, which is the recurring grievance that pushes teams towards self-managed Iceberg stacks in the first place. Iceberg plus object storage plus a serverless query engine is the rational architecture, and the most persuasive detail in the post is that Cloudflare's own billing and infrastructure teams run on it. Query performance at warehouse scale is the question the launch post does not answer, and it is the one that matters.