Deno Sandbox

https://deno.com/deploy/sandbox

Deno Sandbox is a managed service on Deno Deploy that provisions short-lived, isolated Linux microVMs via an API. Each sandbox is API-driven from the @deno/sandbox SDK and is intended for running untrusted or LLM-generated code on behalf of AI agents, CI runners, plug-in systems, or vibe-coding environments. The runtime is a full Linux environment with files, processes, package managers, and background services.

Pricing model Usage-based: CPU while code runs, memory while loaded, plus egress and monthly persistent-volume storage; plan allowances and rates apply
Pricing The headline compute price as the provider states it. Sorting ranks prices stated per vCPU over time, converted to hourly. Other pricing models sort after, unranked. $0.10/CPU-hour; $0.025/GiB-hour of memory; $0.20/GiB-month of volume storage
Free while idle Whether a stopped, paused, or sleeping sandbox costs nothing. ✓
Elastic Resources flex with what the sandbox actually uses, and the bill follows: yes when the provider bills on active or observed use, no when capacity is reserved or allocated and billed while it runs. ✓
Memory snapshots Whether RAM state survives a pause and resume, not just disk. –
Wake on request Whether a stopped sandbox wakes automatically on inbound traffic. –
Start / resume Typical time from create or resume to running. Sorting ranks stated times, converted to seconds; claims with no figure sort after, unranked. under 200 ms (product page demo shows 93 ms)
Max runtime The longest a sandbox may run. 30 min (per docs limits; extendable on demand via extendTimeout)
Isolation The isolation technology between sandboxes. Firecracker microVM
GPUs Whether GPU instances are available for sandboxes. –
Docker Whether Docker containers can run inside the sandbox. –
SDKs TypeScript, JavaScript, Python

✓ yes · ✕ no · – no cited public fact.

Sandboxes are ephemeral by default: files you upload exist only for the sandbox lifetime unless you mount a volume, and the VM is destroyed and the disk wiped when its last reference is dropped or sandbox.kill() is called. Public docs establish filesystem persistence through volumes and read-only snapshots, but do not state that RAM survives pause/resume, and do not establish Docker, GPU, or automatic wake-on-request. Default concurrency during pre-release is 5 sandboxes per organization.

Sources

Last verified 2026-10-01. Corrections land through the update log.