Send2Bucket vs Cloudflare Pipelines
Cloudflare Basin Pipelines, formerly Cloudflare Pipelines, takes HTTP and Workers data, optionally transforms it with SQL, and writes JSON or Parquet files or Iceberg tables to R2 in your Cloudflare account. Send2Bucket writes events and OpenTelemetry data to buckets and catalogs on any cloud.
Short answer: if your data can live in R2, Pipelines is much cheaper per GB and adds SQL transforms. Choose Send2Bucket when you need native OTLP, or when data has to land in S3, Google Cloud Storage, S3 Tables, another Iceberg catalog or DuckLake.
At a glance
| Send2Bucket | Cloudflare Basin Pipelines | |
|---|---|---|
| How you send | POST /v1/events (JSON) and OTLP over HTTP and gRPC | HTTP endpoint, Workers binding, Logpush |
| OpenTelemetry | Native OTLP endpoint | No OTLP; an open-source Worker adapter, otlp2pipeline, exists |
| Transforms | None; records are delivered as sent | SQL transforms |
| Storage | S3, Google Cloud Storage, S3-compatible (including R2) | R2 |
| Tables | Iceberg REST catalogs, S3 Tables, Supabase Analytics Buckets, DuckLake | Iceberg in Basin Catalog |
| Limits | 16 MiB per request after decompression | 5 MB per request, 20 pipelines per account |
| Price | Plans and billing | Ingress free. JSON sink $0.03/GB, Parquet or Iceberg sink $0.06/GB, transforms $0.04/GB, 50 GB of each included. Needs the Workers Paid plan |
When Pipelines fits better
- Your producers already run on Cloudflare Workers.
- You want to filter or reshape records with SQL before they are stored.
- Storing data in R2 and querying it with Cloudflare's tools is fine.
When Send2Bucket fits better
- OpenTelemetry without an adapter. OTLP exporters send straight to Send2Bucket. See the API reference.
- Storage outside Cloudflare. Deliver to Amazon S3, Google Cloud Storage, S3 Tables or any Iceberg REST catalog.
- R2 as one of several targets. R2 works as an S3-compatible destination, so the same API serves R2 and other clouds.