Cloudflare just made its R2 object storage queryable from the dashboard: you can now run SQL directly against R2 Data Catalog tables in a built-in editor, with table exploration, result stats, exports, and EXPLAIN output. That turns a bucket of files into something closer to a data lake you can interrogate without standing up a separate query engine. The strategic point is cost. Cloudflare charges no egress fees on R2, which is exactly where the classic S3-plus-Athena stack quietly bleeds money.

  • A new SQL editor in the Cloudflare dashboard queries R2 Data Catalog tables directly, with exploration, exports, and EXPLAIN.
  • It builds on the open-table approach (Apache Iceberg-style catalogs) so object storage behaves like a queryable lake.
  • R2's zero egress fees are the wedge against S3, where data-transfer and query-scan costs add up.
  • Cloudflare also refreshed its DNS Firewall dashboard, part of a broader push to move power features into the UI.
Querying object storage through R2 Data Catalog Data files sit in R2 object storage, the R2 Data Catalog exposes them as tables, and a SQL editor in the dashboard runs queries directly against those tables and returns results. R2 storage Parquet files Data Catalog tables (Iceberg) SQL editor in dashboard results + EXPLAIN no egress fees on the data you scan genztech.blog
Fig 1 Files in R2 become tables via the Data Catalog, then the dashboard SQL editor queries them in place.

What did Cloudflare actually ship?

A way to query your object storage without leaving the dashboard. R2 Data Catalog exposes files stored in R2 as tables using an open-table format, and the new SQL editor runs interactive queries against them: you can explore tables, see result statistics, export output, and inspect query plans with EXPLAIN. Previously, analyzing data sitting in object storage meant wiring up an external query engine and paying to move or scan the data. Bringing the editor into the console shortens that loop to writing a query and reading the answer.

RelatedEU Finds Instagram, Facebook Addictive Design Breaks DSA

Why does zero egress change the math?

Because in the incumbent model, moving and scanning data is where the bill grows. On the typical S3-based setup, you pay for storage, then pay again to transfer data out and to have a query service scan it. Cloudflare's long-running pitch for R2 is no egress fees, and that advantage compounds for analytics, where the same datasets get queried repeatedly. If querying the data does not rack up transfer charges, the cost of exploratory analysis on large object stores drops meaningfully, which is the whole reason a developer would move a lake to R2 in the first place.

AspectR2 + Data CatalogS3 + external engine
Egress feesNoneCharged
Query surfaceSQL editor in dashboardSeparate service
Table formatOpen (Iceberg-style)Open (Iceberg)
SetupBuilt inProvision + connect

Who is this for?

Teams that already keep logs, events, or analytics data in object storage and want to ask questions of it without a heavyweight warehouse. It is not aiming to replace a full data platform for complex, high-concurrency workloads; it is lowering the barrier for the common case of poking at data you already have. For developers building on Cloudflare's Workers and storage stack, it is another reason to keep everything in one platform rather than shuttling data to a separate cloud to analyze it.

How does this fit Cloudflare's strategy?

It is of a piece with Cloudflare's move to be a full developer platform, not just a CDN and security layer. Adding SQL over R2 and modernizing the DNS Firewall dashboard both push capability into the console so more work happens without leaving Cloudflare. The company keeps chipping at the assumption that serious data and compute must live on AWS, Azure, or Google Cloud. Each feature like this makes the all-Cloudflare stack a little more viable for a little larger class of application.

RelatedChrome's Device-Bound Sessions Go Live to Kill Cookie Theft

What to watch · 2026
  • Query performance. Convenience is set; sustained speed on large scans is the real test.
  • Concurrency limits. Whether this scales past exploratory use into production analytics.
  • Ecosystem hooks. Iceberg-style catalogs mean external tools could query R2 too.
  • Pricing clarity. No egress helps, but compute for queries still needs a transparent model.

How does this pressure the big clouds?

The competitive needle it moves is data gravity, the tendency for workloads to cluster wherever the data already lives. AWS, Azure, and Google Cloud have long relied on egress fees and tightly coupled services to keep data, and therefore compute, on their platforms. Cloudflare attacking that with no-egress storage and now in-place querying chips at the assumption that serious analytics must run on a hyperscaler. It will not flip large enterprises overnight, but for the growing set of developers who start on Cloudflare's edge stack, each feature like this removes another reason to ever reach for a separate cloud to answer a question about data they already hold.

Our take

This is a small feature with a big directional message. Object storage becoming directly queryable, with no egress toll, chips at one of the stickiest reasons data gravity keeps workloads on the big three clouds. Cloudflare is not going to win the data-warehouse category with a dashboard SQL editor, and it does not need to. It needs the all-Cloudflare stack to be good enough that developers stop reflexively reaching for AWS, and every feature that lets you keep and use your data in one place moves that needle. The egress-fee angle is the quiet knockout punch: convenience gets you tried, but a structurally cheaper bill is what gets you kept.

Primary sources

Original analysis by GenZTech. Figures current as of July 2026.