CapabilitiesLifecycle
Pre-started pools
Keep machines ready in advance so new work starts without waiting for a boot.
Short answer: 1 of 16 providers offer it fully: Tensorlake. With a catch: Daytona, Modal, Runtime and Sandboxer. Checked 29 October 2026.
Who offers it
Part of Lifecycle: stop, pause and sleep, which explains why it matters.
Fully: Tensorlake. With a catch: Daytona, Modal, Runtime and Sandboxer. Not at all: Agent 37, Blaxel, Boat, boxd, Cloudflare Sandbox, E2B, InstaBox, Morph, Sprites and Vercel Sandbox.
- TensorlakeYes
A pool keeps a set number of booted containers and refills itself after each claim.
docs-confirmed “A Sandbox Pool is a sandbox template (image, resources, entrypoint, timeout, network policy) plus a set of pre-booted warm containers.” Source
- DaytonaWith a catch
The catch: Pools exist, but they must be switched on for your organisation by asking support.
docs-confirmed “Warm pools must be enabled for your organization. Contact [[email protected]](mailto:[email protected]).” Source
- ModalWith a catch
The catch: You have to build and run the pool yourself; there is no setting that keeps machines waiting.
docs-confirmed “A common use case for this is keeping a pool of Sandboxes available for executing tasks as they come in.” Source
- RuntimeWith a catch
The catch: There is no pool to size or refill; you keep spare paused machines yourself and wake one for each task.
docs-confirmed “For many short tasks, such as fetching one page each, keep spare browsers paused.” Source
- SandboxerWith a catch
The catch: The provider keeps machines ready, but you cannot say how many or which setup; a size the region does not keep boots instead.
docs-confirmed “Each region keeps sandboxes already booted and paused, and a new sandbox from the standard image takes one instead of booting.” Source
- Agent 37No
No pool of pre-started machines; a new instance boots, or is built from a template image.
inferred “Six calls control whether the instance’s computer is running, which image it runs, and how big it is.” Source
- BlaxelNo
No pool to size or fill; machines you already made wait in standby, and a new one boots or resumes from its own snapshot.
inferred “Sandboxes resume from standby in under 25 milliseconds and automatically scale to zero after a few seconds of inactivity.” Source
- BoatNo
No pool of pre-started machines.
inferred “A resume, a fork or a deploy is ready in **a few seconds, whatever the size of the disk**.” Source
- boxdNo
No pool of pre-started machines; every machine is created or woken on demand.
inferred “A machine moves between these states on its own, based on network traffic.” Source
No pool of pre-started machines; a start creates a new instance.
inferred “When work arrives, the Durable Object calls [`this.ctx.container.start()`](/containers/api/durable-object-container/#start), which starts a container instance” Source
- E2BNo
No pool of pre-started sandboxes.
inferred “The sandbox lifecycle: create, connect to, pause, resume, and kill an isolated Linux VM for your agent, and control what happens when its timeout expires.” Source
- InstaBoxNo
No pool of pre-started machines; the paid plans sell a dedicated pool only as an enterprise option.
inferred “Dedicated pool or BYOC” Source
- MorphNo
No pool of pre-started machines; branch makes copies on demand.
inferred “Each instance runs from a saved snapshot.” Source
- SpritesNo
No pool of pre-started machines; each Sprite starts and pauses on its own.
inferred “A Sprite is not a server that runs until you turn it off.” Source
No pool of pre-started machines.
inferred “Each time you resume a sandbox, a new session starts from the last saved state.” Source
- LizardNot published
Nothing documents a pool you can create; the docs only mention pool capacity in passing.
not-published “Availability depends on the region and pool capacity.” Source
Related
Need several capabilities at once? Add more in the capability finder. Something wrong here? Use the report link at the bottom of the page.