#SandBoxes

An AI coding agent can read a repository, run shell commands, install packages, and send network...

You can now run uninterrupted sessions on Vercel Sandbox for up to 24 hours supporting very long running agents

Use any image from VCR with Vercel Sandbox to provide a custom base filesystem for your Vercel Sandbox allowing you to bring custom tools to your Sandbox

Vercel Sandbox now has full support in Vercel Observability, giving you visibility into how your sandboxes consume resources.

You can now open an interactive terminal to any Sandbox: open a shell into a running sandbox, browse the filesystem, run commands, upload and download files, open ports, and much more

The Vercel Sandbox CLI now ends create output with a connect hint, keeps sandbox not found errors quiet and actionable, explains the right command when sh is misused, and warns when the installed CLI is outdated.

Connect Vercel Sandboxes to a dedicated Secure Compute network for static outbound IPs and access to private AWS resources through VPC peering.

Self-hosted dev sandboxes with preview URLs. One command. No Kubernetes, perfect for coding agents and Saas factories - tastyeffectco/sandboxes

Run up to 2000 sandboxes at the same time. Sandbox also now supports listening on up to 4 different ports

The Sandbox CLI, based on the Docker CLI, allows you to manage sandboxes, execute commands, copy files, and more from the command line.

Vercel Sandbox is now GA, providing isolated Linux VMs for safely executing untrusted code. Run AI agent output, user uploads, and third-party code without exposing production systems.

Run Vercel Sandbox commands directly from the Vercel CLI with vercel sandbox, without needing to install the standalone Sandbox CLI.

Tags on Vercel Sandbox is now in open beta, add tags to a sandbox, update tags for an existing sandbox and filter by tags when listing sandboxes.