Agent sandboxes / Leap0 alternative

A Leap0 alternative for agents that accumulate system state

Leap0 already provides Firecracker machines with both memory-preserving pause and disk-preserving stop. Its snapshots capture writable disk and execution state into reusable checkpoints. For a returning agent, the main questions are storage growth, timeout policy, and which requests wake compute.

Start buildingCompare capabilities ↓

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

When Leap0 fits

Leap0 fits teams that want its Linux microVM, desktop API, persistent interpreter, Git operations, LSP, and credential-brokering network transforms. Its named snapshots can prepare or branch an execution environment rather than reinstall the same tools for every run.

When to consider Sandbox0

Consider Sandbox0 when the complete durable writable RootFS should grow online across tasks and supported file operations should work without waking compute. Check desktop, full-kernel, interpreter, and tool API requirements separately.

Capabilities

Leap0 and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeLeap0The documented preview timeout supports up to eight hours; this is a bounded session setting, not indefinite runtime.View evidence ↗Sandbox0ttl 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 filesLeap0Stop/start keeps writable disk changes; snapshots are separately named reusable checkpoints.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSLeap0Snapshots capture writable disk and memory, and stop preserves writable disk rather than only a workspace folder.View evidence ↗Sandbox0Persistence extends beyond /workspace to paths such as /usr/local, /etc, and /home. Runtime tmpfs and virtual filesystems are excluded.Read documentation →
Release compute, keep stateLeap0Pause retains execution state and automatically resumes on a request; stop/start is filesystem-only.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateLeap0Snapshots explicitly include in-memory execution state and writable disk.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.

The reviewed limits specify a 10 GiB disk; they do not establish automatic online RootFS growth. Sandbox0’s enabled imports grow the existing mounted filesystem within its configured sparse-device ceiling.

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.

Leap0’s paused lifecycle wakes on a request, including its runtime tool surface. The reviewed docs do not establish file access that leaves paused compute stopped.

See paused file operations →

Pause, stop, and snapshot restore have different identities

A paused Leap0 sandbox has saved memory and disk and wakes on a request. Stop keeps writable disk but discards in-memory execution; start boots that same stopped sandbox. Restoring a named snapshot instead creates a new sandbox, taking its compute settings from the checkpoint. A client should preserve the identifier matching its chosen workflow.

Source 1 ↗

Current storage and timeout limits bound a returning environment

The preview limits list a 10 GiB disk and a timeout of one minute to eight hours, defaulting to five minutes. Auto-pause can park at the timeout. That is different from an online-growing filesystem or a promise of an unbounded active session; confirm preview quotas as they change.

Source 1 ↗Source 2 ↗

Preview access and future unit rates are separate offers

The homepage says access is free during public preview and the displayed CPU/RAM rates apply when billing launches. Those future rates are useful planning inputs, but they are not the present paid invoice. The preview guide also warns of breaking changes and downtime.

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.

Leap0

Pricing reviewed 2026-10-18

Current public preview
Free; no card required
Rate source ↗
Announced CPU rate after billing launches
$0.0504 per vCPU-hour
Future paid rate, not a current preview charge.
Rate source ↗
Announced RAM rate after billing launches
$0.0162 per GB-hour
Preserve the published GB unit.
Rate source ↗

The current offer is free preview. A complete future retained-disk/snapshot invoice is not established by the displayed CPU/RAM table. Compare production requirements and effective quotas separately from vendor startup-latency claims.

Check current Leap0 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 Leap0

  1. Inventory pause versus stop, named snapshots, timeout/auto-pause settings, system packages, interpreter state, desktop dependencies, and actual retained files.
  2. Export required data and port compatible images, commands, files, credentials, networking, and any separate agent tools.
  3. Validate complete system-path retention, disk growth, paused file access, process-continuity requirements, and effective post-preview billing.

Common questions

Is Sandbox0 a drop-in Leap0 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 ↗leap0.dev/docs/llms-full.txt ↗leap0.dev/docs/limits ↗leap0.dev/ ↗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 →