Agent sandboxes / AWS Lambda MicroVMs alternative

AWS Lambda MicroVMs alternative

Lambda MicroVMs is a distinct AWS compute product for directly managed isolated environments, rather than the ordinary event-function invocation model. It supports Dockerfile-built images, dedicated HTTPS endpoints and memory/disk suspend-resume. Compare the retention window, filesystem growth and full snapshot bill for a long-lived agent.

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

When AWS Lambda MicroVMs fits

Choose Lambda MicroVMs when Firecracker isolation, AWS networking, snapshot-based startup and automatically bursting compute fit the workload. Its retained memory and disk are real capabilities, not merely a workspace-volume feature.

When to consider Sandbox0

Consider Sandbox0 for a durable Linux runtime identity whose complete writable RootFS can grow online, with supported file access while compute remains paused. Assess architecture, network, SDK and explicit memory-checkpoint compatibility before porting the workload.

Capabilities

AWS Lambda MicroVMs and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeAWS Lambda MicroVMsThe documented MicroVM workflow has an eight-hour session boundary and bounded suspension. This is distinct from the fifteen-minute Event Function limit.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 filesAWS Lambda MicroVMsSuspension saves disk state for the supported retention window.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSAWS Lambda MicroVMsMicroVM snapshots preserve application memory and disk, including its initialized runtime filesystem.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 stateAWS Lambda MicroVMsAutomatic and programmatic suspend/resume are documented.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateAWS Lambda MicroVMsMemory and disk state are preserved during bounded suspension.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.

Lambda MicroVMs already preserve memory and disk through supported suspension. Automatic compute bursting is acknowledged, while automatic online growth of the complete durable RootFS is not established by the reviewed MicroVM contracts.

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.

The reviewed guides describe accessing the running MicroVM through its endpoint and resuming suspended state. They do not establish direct full-RootFS file mutations while compute remains suspended; snapshot storage is a different interface.

See paused file operations →

Suspension preserves memory and disk with a time bound

AWS documents memory/disk snapshots and up to eight hours of suspension without compute charges. The launch workflow also describes sessions spanning eight hours. This is different from an indefinitely retained environment; set expiry and export policies around the actual service limits.

Bursting compute is separate from growing storage

A configured baseline allocates CPU in a 2:1 memory-to-CPU ratio. Running baseline resources are billed even when not fully consumed, while additional burst resources are charged for active consumption. Automatic CPU/memory bursts do not establish automatic expansion of a durable RootFS.

Each suspend, resume and launch can add snapshot charges

Snapshots are billed for storage, data written on suspension and data read on launch/resume. Image storage has a one-week minimum retention period. Frequent idle cycles can therefore change the bill even while suspended compute is 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.

AWS Lambda MicroVMs

Pricing reviewed 2026-10-19

ARM vCPU, US East N. Virginia reference
$0.0000276944 per vCPU-second
Published MicroVM example; baseline plus active burst rules apply.
Rate source ↗
ARM memory, same reference region
$0.0000036667 per GB-second
Separate from CPU; preserve AWS GB units.
Rate source ↗
Snapshot data written
$0.0038 per GB
Reference example charges this on suspend.
Rate source ↗
Snapshot data read
$0.00155 per GB
Reference example charges launch and resume reads.
Rate source ↗
Snapshot storage
$0.08 per GB-month
Reference region/architecture example; image storage has a one-week minimum.
Rate source ↗

Rates above are the published US East (N. Virginia), ARM MicroVM example, not a global price guarantee or an Event Functions tariff. Snapshot sizes, cycle count, image retention, network transfer and VPC connectivity affect the total.

Check current AWS Lambda MicroVMs 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 AWS Lambda MicroVMs

  1. Inventory images, ARM compatibility, baseline/burst resources, maximum duration, idle policy, suspension window and snapshot cycle volume.
  2. Export essential filesystem state and port application endpoints, credentials and network requirements into a compatible Sandbox0 template.
  3. Test system-tool retention, online disk growth, required process restoration, paused files and total compute/snapshot/network charges.

Common questions

Is Sandbox0 a drop-in AWS Lambda MicroVMs 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