Agent sandboxes / Depot sandboxes alternative

Depot.dev alternative for persistent agent sandboxes

Depot offers CI/build infrastructure, remote Claude Code agent sandboxes and a separate programmatic Sandbox SDK. Remote agent sessions save filesystem and conversation state; the public-beta SDK exposes command and file operations with a bounded lifetime. Compare the product you actually integrate, because these surfaces have different billing and recovery contracts.

Start buildingCompare capabilities ↓

Reviewed 2026-10-19 · Published by the Sandbox0 team

When Depot sandboxes fits

Choose Depot when remote Claude Code collaboration or its build, cache and CI infrastructure are central. Its agent-session filesystem saving is a real persistence feature; preserve that session/harness behavior when you depend on it.

When to consider Sandbox0

Consider Sandbox0 for a stable Linux runtime identity, whole-RootFS online growth and supported paused-file operations. You supply the agent harness, shared conversation state and any Depot build/cache integrations you retain.

Capabilities

Depot sandboxes and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeDepot sandboxesThe beta SDK defaults to 120 minutes from creation, including provisioning, and caps lifetime at twenty-four hours. Extension cannot pass that absolute maximum; remote-agent sessions are a separate product.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 filesDepot sandboxesRemote agent sessions save files for later session-ID recovery; this is distinct from the SDK lifecycle.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSDepot sandboxesFilesystem persistence is explicit, but complete writable RootFS path coverage is not established.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 stateDepot sandboxesAgent-session recovery and SDK timeout/terminal states are distinct; a uniform machine-pause contract is not established.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateDepot sandboxesSaved conversation context does not establish arbitrary process-memory recovery.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.

Depot remote agents already save filesystem state. Its SDK permits an initial disk size, but the reviewed contracts do not establish automatic online expansion of a whole durable RootFS or common retention across every Depot execution product.

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.

SDK file operations are documented on active sandbox handles, and remote agents load saved state on resumption. The reviewed guides do not establish whole-RootFS file operations while sandbox compute remains stopped.

See paused file operations →

Remote agent sessions save filesystem and conversation state

The agent guide saves files and context when the agent exits and resumes them through a session ID. This is useful returning-agent state. The guide does not define complete writable system-path coverage, an arbitrary RAM checkpoint or automatic online RootFS growth.

Source 1 ↗

The beta SDK is a distinct sandbox lifecycle

The SDK defaults to a 120-minute lifetime from creation, including provisioning, and caps lifetime at 24 hours. setTimeout can extend a running deadline only within the absolute maximum. Stop/kill end the sandbox; the reviewed SDK does not document importing remote Claude session recovery into this API.

Source 1 ↗

Do not combine two different sandbox rate cards

Remote agents list $0.01/minute with no included agent usage. The SDK instead states billing at the Depot CI per-vCPU-second rate. The public price page lists CI at $0.00005/vCPU-second and a separate agent minute meter. Obtain the effective product terms rather than applying the flat agent rate to every SDK resource shape.

Source 1 ↗Source 2 ↗Source 3 ↗

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.

Depot sandboxes

Pricing reviewed 2026-10-19

Developer platform
$20/month
One user; separate build/CI/storage allowances do not include remote agent minutes.
Rate source ↗
Remote agent sandboxes
$0.01 per minute
No included remote-agent usage; default agent shape documented as 2 vCPU/4 GB.
Rate source ↗
Beta SDK / Depot CI compute
$0.00005 per vCPU-second
SDK uses the CI meter; resource count changes cost. Confirm effective beta/account terms.
Rate source ↗
Startup platform
$200/month
Unlimited users and larger listed CI/build/storage allowances; agent and SDK usage remain separately scoped.
Rate source ↗

SDK, Depot CI, GitHub Actions runners, container builds and remote agents are separate meters. Filesystem recovery in the agent guide does not establish durable RAM or SDK stop/resume. Preserve the beta qualification, absolute lifetime and actual allocated resource units.

Check current Depot sandboxes 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 Depot sandboxes

  1. Identify remote agent sessions versus beta SDK/CI, capture session state, resource shapes, expiry and effective rate terms.
  2. Export essential files and integrate the desired agent harness with Sandbox0; preserve build and collaboration services separately.
  3. Test system-path retention, process expectations, online growth, paused files and all platform, per-vCPU, agent and model charges.

Common questions

Is Sandbox0 a drop-in Depot sandboxes 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 ↗depot.dev/docs/agents/overview ↗depot.dev/docs/api/sandbox-sdk-reference ↗depot.dev/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 →