Agent sandboxes / Runta alternative

A Runta sandbox alternative for accumulated agent files and tools

Runta combines agent execution with memory scaling, checkpoint restore, and governed network access. Its memory can grow without recreating a runtime, but memory growth is a separate property from the capacity of an agent’s persistent root filesystem.

Start buildingCompare capabilities ↓

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

When Runta fits

Runta fits teams that want its agent execution layer, full-state checkpoints, optional memory autoscaling, and suspension around published services. Those features can support agent workloads that vary in memory demand or spend time waiting.

When to consider Sandbox0

Consider Sandbox0 when accumulated system packages, configuration, user files, and outputs should share an automatically growing durable RootFS. Supported persistent-file operations during pause help the application manage that state without waking compute.

Capabilities

Runta and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeRuntaNot 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 filesRuntaCheckpoints retain filesystem state for later restore into a new runtime.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSRuntaA checkpoint is described as capturing runtime filesystem and running 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 stateRuntaThe preview idle-suspend workflow restores a runtime and proxies incoming traffic.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateRuntaCheckpoint state includes running processes, rather than only a prepared workspace.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.

Runta’s reviewed autoscaling guide concerns RAM, not mounted-root filesystem capacity. Its checkpoint and memory APIs do not establish automatic free-space-triggered growth 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.

See paused file operations →

Memory autoscaling has an explicit ceiling

Runta scales memory between an initial allocation and a configured maximum. Omitting the maximum disables growth. Its guide also qualifies the feature for gradual increases rather than sudden spikes. That is useful capacity management, but it does not establish root-disk autoscaling.

Source 1 ↗

A checkpoint restore creates another runtime

Runta checkpoints capture filesystem state and running processes. Restoring creates a new runtime, so retain the checkpoint ID and reconnect the application to the restored machine. A reusable full-state capture and a continuously reused sandbox identity are distinct lifecycle choices.

Source 1 ↗

Idle suspension and storage have their own meters

Runta’s auto-suspend and wake-up guide is marked preview and measures inactivity using inbound HTTP traffic. A later idle scan performs suspension, so its timeout is not a precise deadline. The price page stops CPU and memory charges while paused, suspended, or shut down, but retains disk-storage charges.

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.

Runta

Pricing reviewed 2026-10-18

Running CPU
$0.0504 per vCPU-hour
Rate source ↗
Running memory
$0.0162 per GiB-hour
Rate source ↗
Retained disk storage
$0.000108 per GiB-hour
Disk-only billing while paused, suspended, or shut down; deleting the runtime ends these listed state charges.
Rate source ↗

Runta lists prepaid usage with no subscription or monthly minimum, a $50 trial balance, and a $5 minimum manual top-up. Resource allocation, memory-scaling configuration, retained artifacts, and model or agent-layer services should be checked for the intended workload.

Check current Runta 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 Runta

  1. Inventory checkpoint IDs, current runtime identity, memory bounds, idle-suspension policy, and published services before teardown.
  2. Port the image, command and file calls, credentials, egress policies, and service startup; reconnect clients when runtime identity changes.
  3. Validate growing files and installed tools across tasks, paused file access, required process restoration, and active versus retained-storage costs.

Common questions

Is Sandbox0 a drop-in Runta 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 ↗runta.com/docs/runtime/memory-auto-scaling/ ↗runta.com/docs/runtime/checkpoints/ ↗runta.com/docs/runtime/auto-suspend-and-wake-up/ ↗runta.com/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 →