Cloudflare Containers and Sandboxes are finally GA!

Hey everyone, it's me, Shii-chan! Today I've got a "it's finally the official release!" kind of news that folks who've been waiting will love.

Cloudflare Changelog developers.cloudflare.com

What was announced?

On the Cloudflare Changelog, Containers and Sandboxes were announced as generally available (GA). Containers let you run CLI tools that need a full Linux environment, apps in many languages, and heavier workloads right next to your Workers. It's on the Workers Paid plan.

The story so far

Until now, Containers had been steadily stacking up features since launch. Going GA means it's no longer "try it out and stop there" — it's Cloudflare's sign-off that you can rely on it in production. It reached the official release with all the post-launch improvements baked in.

What changes

The big deal is that you can now more seriously pick it as a base for running containers. Thanks to the higher resource limits added before GA, it can handle thousands of concurrent containers, and billing is an Active-CPU model that only charges for the CPU cycles you consume. Not paying for idle time is great news if you want to keep containers around for a while.

Dive Deep

Here are the improvements that landed on the Containers side after launch:

• Active-CPU pricing, so you only pay for the CPU you consume
• Easy connections to Workers and other bindings via hostnames
• Use existing images and registries with Docker Hub
• SSH access for debugging

And Sandboxes (the Sandbox SDK) gained a bunch too: live preview URLs to check long-lived services, persistent code interpreters for Python, JavaScript, and TypeScript, browser-based interactive PTY terminals, backup and restore APIs for snapshotting your workspace, and even real-time filesystem watching that reacts instantly to file changes. Handy when you want to embed a whole code-execution environment.

Wrap-up

• Cloudflare Containers and Sandboxes are now GA (generally available)
• It's on the Workers Paid plan; the highlights are thousands of concurrent containers and Active-CPU pricing
• Docker Hub integration, SSH, preview URLs, PTY terminals, backup/restore, and more — all the post-launch improvements are on board

If you want to run a full Linux environment near the edge, or safely let an AI agent execute code, this is your cue to start thinking about production!