Agent sandboxes / Sail Research (Sailboxes) alternative

A Sail Research alternative for a growing agent computer

Sailboxes already checkpoints the whole machine when it sleeps, including disk, memory, and processes. The useful comparison is its observed-usage billing, configured storage ceiling, independent volumes, and wake triggers versus Sandbox0’s automatically growing durable RootFS.

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

When Sail Research (Sailboxes) fits

Sailboxes fits teams that want reusable agent machines beside Sail inference or training workflows, memory-preserving sleep, scheduled commands, and billing based on observed CPU, RAM, and disk usage rather than reserved ceilings.

When to consider Sandbox0

Consider Sandbox0 when the runtime should supply a per-user durable RootFS that grows automatically while the agent works, with supported file reads and filesystem-only edits available while compute remains paused.

Capabilities

Sail Research (Sailboxes) and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeSail Research (Sailboxes)Not 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 filesSail Research (Sailboxes)Sleeping restores disk state; independent volumes have their own lifetime.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSSail Research (Sailboxes)Sleep checkpoints the whole machine, including disk and processes.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 stateSail Research (Sailboxes)Sleep and pause retain state, with different wake triggers.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateSail Research (Sailboxes)Memory and running processes return after sleep.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.

Sailboxes exposes create-time disk ceilings by size, with documented bounds through 1 TiB on L. Observed disk-usage billing and full-state sleep do not by themselves establish automatic online expansion of the complete durable RootFS.

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.

Sail lists file transfers among the operations that wake a sleeping Sailbox. Sandbox0’s supported paused Files API explicitly leaves compute paused, with mutations restricted to filesystem-only pauses.

See paused file operations →

Sleep retains the whole machine, while pause controls waking

Sail’s autosleep guide explicitly restores disk, memory, and running processes. Sleeping machines wake on exposed traffic, commands, file transfer, SSH, or scheduled work. Manual pause also captures state but ignores traffic-based wake, so choose the lifecycle that matches background services and inspection.

Resource ceilings are not reserved usage

A Sailbox size sets CPU and default memory and disk ceilings, with configurable bounds at creation. The bill tracks observed usage below those ceilings. Distinguish the disk limit from actual data consumption and from a policy that grows a mounted root filesystem automatically.

Sleeping compute and retained volumes use separate billing

Sail does not bill running resources during sleep, pause, checkpointing, or cold start. Independent volume storage is still billed for each hour the volume exists. Free also has size-based creation charges, while Pro and Enterprise waive those creation fees rather than making resource usage free.

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.

Sail Research (Sailboxes)

Pricing reviewed 2026-10-18

Observed running CPU
$0.015 per used vCPU-hour
Rate source ↗
Observed running RAM
$0.008 per used GiB-hour
Rate source ↗
Observed running NVMe disk
$0.0007 per used GiB-hour
Rate source ↗
Independent volume storage
$0.000411 per GiB-hour
Billed while the volume exists, including when attached Sailboxes sleep.
Rate source ↗
Free-plan S / M / L creation
$0.005 / $0.01 / $0.012 per creation
Waived on Pro and Enterprise.
Rate source ↗
Pro plan
$250 per month
Includes $100 monthly credits; the current page adds $150 in the first month.
Rate source ↗

Running usage is measured every second and added to the bill in one-minute increments. Plan credits, creation charges, independent volumes, and inference usage remain separate considerations. The published size table reaches a configurable disk ceiling of 1 TiB on L; this is a ceiling, not evidence of automatic mounted-filesystem growth.

Check current Sail Research (Sailboxes) 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 Sail Research (Sailboxes)

  1. Inventory Sailbox state, independent volumes, size ceilings, sleep/pause choices, exposed services, and scheduled work.
  2. Export the required files and port images, commands, model integration, credentials, and wake policy; reconnect services after process restoration.
  3. Validate growing system packages and outputs, whole-RootFS continuity, paused files, and observed usage versus Sandbox0’s selected memory tier.

Common questions

Is Sandbox0 a drop-in Sail Research (Sailboxes) 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.

Related comparisons