Pricing

Pipelines charges based on two dimensions:

1. SQL transforms: The volume of data processed by stateless SQL.
2. Sinks: The volume of data delivered to each sink destination.

Ingress into a Pipeline stream is free. Standard R2 storage and operations charges apply for data written to R2 buckets. R2 Data Catalog charges apply when writing to Iceberg tables.

All included usage is on a monthly basis.

Streams provide durable, distributed log storage that buffers incoming messages. Ingress into a stream is free regardless of volume. A single stream can be read by multiple pipelines.

SQL transforms let you filter, reshape, and compute over data before it reaches a sink. Any query currently counts as a transform.

Pricing covers stateless transforms (for example, filter, reshape, unnest, cast, and compute). Future stateful operations such as aggregations, joins, and windows may be priced separately.

Sink pricing is based on the volume of uncompressed data delivered to the destination. The rate varies by output format:

• JSON: $0.03 / GB — lowest compute cost, suitable for simple log forwarding.
• Parquet / Iceberg: $0.06 / GB — higher compute cost for columnar encoding and Iceberg table management. Best for analytics workloads.

A pipeline ingests 500 GB of event data per month. A SQL transform filters and reshapes the data, reducing output to 300 GB written to an R2 Data Catalog Iceberg table.

To learn more about how usage is billed, refer to Cloudflare Billing Policy.