Runloop is winding down. Its documentation lists existing-customer support through November 13, 2026 at 11:59 PM PT. Review this deadline before planning a migration. Read the source ↗

Agent sandboxes / Runloop alternative

A Runloop Devbox alternative for keeping agent environments

Runloop’s documentation announces a service wind-down with support for existing customers through November 13, 2026 at 11:59 PM PT. Prioritize exporting environments and replacing orchestration, then validate persistence and cost on the destination.

Start buildingCompare capabilities ↓

Reviewed 2026-10-18 · Published by the Sandbox0 team

When Runloop fits

Runloop remains relevant to existing customers finishing a transition within the published support window. Its Devbox and blueprint APIs describe the source environment to export; confirm service access directly before planning any new dependency.

When to consider Sandbox0

Consider Sandbox0 as a destination for persistent Linux agent environments. Keep installed tools and user state in the writable RootFS, configure retention, and use pause/resume to release compute between tasks.

Capabilities

Runloop and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeRunloopNot established in this reviewSandbox0ttl and hard_ttl default to 0 (disabled). Configure idle pause and deletion policies for the same persistent sandbox identity; an idle running sandbox still accrues compute charges.Read documentation →
Persistent filesRunloopSuspend saves Devbox disk state; retain snapshots during transition.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSRunloopNot established in this reviewSandbox0Persistence extends beyond /workspace to paths such as /usr/local, /etc, and /home. Runtime tmpfs and virtual filesystems are excluded.Read documentation →
Release compute, keep stateRunloopSuspend/resume releases compute and restores disk state.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateRunloopThe documented suspend/resume path does not preserve in-memory state.View evidence ↗Sandbox0Experimental memory checkpoints can preserve compatible process state. Use memory: true on both pause and resume on a compatible deployment. A filesystem checkpoint alone does not preserve running processes.Read documentation →

Online RootFS growth

On growth-enabled images, Sandbox0 automatically expands the writable RootFS as free space runs low, within a configured sparse logical ceiling up to 64 TiB. Agent processes keep running. Growth is enabled in the qualified Cloud region; confirm activation for other regions. Existing images and snapshots retain their captured storage geometry. The ceiling is logical capacity, not preallocated host disk.

For migration, record the actual source disk size and used data. Validate the destination’s automatic RootFS growth with the exported workload instead of assuming a Devbox disk specification transfers unchanged.

Files while compute is paused

Sandbox0's Files API reads persistent files without resuming compute. Filesystem-only pauses also permit edits; retained-memory pauses are read-only and reject mutations with 409. SSH, SFTP, file watching, and runtime-only mounts require running compute.

See paused file operations →

Treat this as an environment migration

The official documentation now carries a wind-down notice. Capture reusable images, repositories, secrets references, and data while access remains available. Verify the destination with a representative task before removing the source.

Source 1 ↗

Devbox suspend preserves disk rather than RAM

Runloop documents suspension as a disk snapshot workflow and explicitly excludes in-memory state. Rebuild background services and interpreters on resume. Sandbox0’s default also starts fresh processes; compatible memory restoration is an explicit experimental option.

Source 1 ↗Source 2 ↗

Keep the historical bill useful for planning

The published Pro price is $250/month plus separately billed usage, with different meters for running Devbox disks and snapshot storage. These rates describe the source bill; the wind-down notice means they should not be treated as a guarantee of future service availability.

Source 1 ↗Source 2 ↗

Pricing

Compare billing models, then price your workload

CPU allocation, memory, billing minimums, disk, plan fees, and supported workload differ. These meters are not equal-performance machine quotes. Confirm region, resource units, and included allowances before estimating a total.

Runloop

Pricing reviewed 2026-10-18

Published Devbox CPU
$0.108 per CPU-hour
Rate source ↗
Published Devbox memory
$0.0252 per GB-hour
Rate source ↗
Published Devbox storage
$0.00034236 per GB-hour
Rate source ↗
Published snapshot storage
$0.000072 per GB-hour
Rate source ↗
Published Pro plan
$250/month plus usage
Rate source ↗

These are the source service’s published rates during wind-down, not a forward availability commitment. Usage is additional to the Pro subscription; suspended environments still retain billable storage.

Check current Runloop pricing ↗

Sandbox0

$0.015/GiB-hour compute (billed by the second at approximately $0.000004167/GiB-second), plus $0.02/GiB-month of retained incremental RootFS and memory checkpoint data. Shared base images and platform recovery retention are excluded; paused sandboxes accrue storage charges only.

Compute / GiB-hour
$0.015
Retained data / GiB-month
$0.02
Paused compute
$0.00

A 4 GiB agent, 100 running hours

$6.00 compute

4 GiB × 100 hours, billed by the second. Keeping 20 GiB of billable incremental data for a full billing month adds $0.40. CPU is derived from memory by the platform; choose a tier that fits your workload.

Understand storage billing →

Before migrating from Runloop

  1. Export Devbox data, snapshots, blueprints, object-storage inputs, and configuration before the published support window ends.
  2. Rebuild source images and replace Runloop named-shell, process, credential, and tunnel integrations.
  3. Run the agent on Sandbox0, verify filesystem checkpoints and paused file access, and confirm retention before retiring the source.

Common questions

Is Sandbox0 a drop-in Runloop replacement?

Evaluate API compatibility and product scope. A shared sandbox category does not make SDKs, lifecycle semantics, or limits interchangeable.

Does keeping files also keep a running agent?

Filesystem state and volatile process state are separate. Sandbox0 requires explicit compatible memory checkpoints to preserve running processes; its default filesystem resume starts fresh processes.

Sources and review scope

The Boat team's Battleships directory is the discovery source. References below describe the product and plan discussed on this page. Check current availability, regional terms and billing units before migrating. Where the cited material does not establish a capability or price, the comparison identifies that uncertainty.

Battleships research record ↗docs.runloop.ai/docs/devboxes/lifecycle ↗/docs/sandbox/pause-resume ↗www.runloop.ai/pricing ↗Sandbox0 RootFS capacity →

Related comparisons

Agent 37 vs Sandbox0 →AgentComputer vs Sandbox0 →ainclave vs Sandbox0 →Alibaba Cloud Agent Sandbox / FC / AgentRun vs Sandbox0 →Alibaba Cloud AgentBay vs Sandbox0 →Amika vs Sandbox0 →