NewsCloud & DevOpsInfrastructure

Cloudflare Basin Is GA: Serverless Analytics, Zero Egress Fees

Cloudflare Basin serverless analytics platform with Apache Iceberg and zero egress fees

Cloudflare Basin went generally available on October 1. It is a serverless data analytics platform built on Apache Iceberg and R2 object storage, and its headline feature is deceptively simple: you do not pay to move your data out. That single line is a direct attack on a billing pattern that has quietly drained engineering budgets for a decade.

Three Products, One Pipeline

Basin is not a monolith. It ships as three coordinated products covering every stage of an analytics workflow.

Basin Pipelines handles ingestion. It accepts events from Cloudflare Workers, HTTP endpoints, and Cloudflare Logpush, applies SQL transforms in flight, and writes to Apache Iceberg tables in R2. Ingress is free at any volume. The pipeline sustains up to 3 GB per second per stream — enough for high-throughput production workloads.

Basin Catalog manages the metadata layer. It automatically maintains your Iceberg tables: compacting small files, expiring old snapshots, and keeping table statistics fresh so queries stay fast without manual maintenance. Any Iceberg-compatible engine — DuckDB, Trino, Spark, Snowflake — can read the same tables directly. No copy, no conversion.

Basin SQL is the query engine. It is serverless and distributed, supports table joins, window functions, and over 190 SQL functions. You send a query, you get results. No cluster to size, no auto-scaling policies to tune.

Why Egress Fees Matter

AWS charges $0.09 per GB to transfer data out to the internet. Athena itself does not add an egress line item, but every time you pull query results from S3 to your application or BI tool, that $0.09 applies. At modest scale it is invisible. At real scale it is a surprise on the invoice that grows faster than your query volume.

Basin’s answer is structural: your data lives in R2, your queries run in Basin SQL, and results stay within Cloudflare’s network. There is no billing boundary to cross. Egress is zero by architecture, not by promotional pricing that expires.

Apache Iceberg Is the Escape Hatch

The legitimate concern when any cloud vendor launches a data platform is whether you are trading one lock-in for another. Basin addresses this with the storage format itself. Data in Basin Catalog is Apache Iceberg — an open standard under the Apache Software Foundation, not owned by Cloudflare or anyone else.

Spark, Trino, Flink, DuckDB, Snowflake, and BigQuery all read Iceberg natively. If you want to switch query engines, you point the new engine at the existing catalog. No migration job, no reformatting, no renegotiating contracts. The data stays where it is; the compute is swappable.

Already Running in Production

Anomaly — the company behind open-source AI coding agent OpenCode — replaced their entire AWS S3 and Athena data pipeline with Basin before the GA announcement. Co-founder Dax Raad described the result as “a cleaner, serverless architecture that reliably handles all of our event data.” That is not a benchmark on synthetic data; it is a production workload migration completed before general availability.

How to Start

The entry path is straightforward if you are already on the Workers platform. Add a pipeline binding in your wrangler.toml and send events from your Worker:

// wrangler.toml
[[pipelines]]
binding = "ANALYTICS_PIPELINE"
pipeline = "my-analytics"

// worker.js
await env.ANALYTICS_PIPELINE.send([{
  path: new URL(request.url).pathname,
  timestamp: Date.now()
}]);

Workers Free includes 1 GB per month. Workers Paid ($5/month) includes 50 GB. Ingress is always free. Transforms cost $0.04 per GB beyond the included tier. Full Basin documentation and getting started guides are on the Cloudflare developer site.

The Bigger Picture

Basin is the third significant infrastructure product Cloudflare has shipped in short succession: K2 replaced managed Kafka brokers, Clef added decision-layer logic for AI agents, and Basin now targets the analytics data stack. The pattern is consistent — take a component that traditionally requires a dedicated managed service, strip out the operational overhead, and undercut on egress. Cloudflare is not competing at the CDN margin anymore.

Basin at GA is a credible option for new analytics workloads, especially for teams already invested in the Workers ecosystem. The open-format foundation means you are not betting on Cloudflare remaining the right choice forever — you are just betting that Apache Iceberg will remain readable, which is a much safer bet. For workloads still on AWS S3 and Athena, the egress math is worth running before your next renewal cycle.

ByteBot
I am a playful and cute mascot inspired by computer programming. I have a rectangular body with a smiling face and buttons for eyes. My mission is to cover latest tech news, controversies, and summarizing them into byte-sized and easily digestible information.

    You may also like

    Leave a reply

    Your email address will not be published. Required fields are marked *

    More in:News