Computer-use platforms / Windows 365 for Agents alternative

Windows 365 for Agents alternative for persistent agent state

Windows 365 for Agents provides pooled Cloud PCs for agent execution with Microsoft identity and management integrations. Its security guide explicitly resets the machine after every agent session with no state carried forward. Always-available readiness is therefore different from retaining one agent’s evolving installed computer.

Start buildingCompare capabilities ↓

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

When Windows 365 for Agents fits

Choose Windows 365 for Agents for supported Microsoft agent orchestration, enterprise identity/policy and isolated task-oriented Cloud PC execution. The supported integration and guest OS determine which applications and agent workflows it can run.

When to consider Sandbox0

Consider Sandbox0 for compatible Linux agents that should return to the same durable whole writable RootFS, grow it online and use supported files during compute pauses. Keep required native Windows applications and Microsoft orchestration services separately.

Capabilities

Windows 365 for Agents and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeWindows 365 for AgentsNot 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 filesWindows 365 for AgentsThe documented pooled Cloud PC is reset after every agent session with no state carried forward.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSWindows 365 for AgentsThe stateless per-session model does not retain the prior agent’s arbitrary writable OS changes.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 stateWindows 365 for AgentsAlways-available readiness is not a documented memory-preserving returning-agent pause/resume contract.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateWindows 365 for AgentsStateless session pooling does not establish user-controlled live-process/RAM snapshot 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.

The agent pool resets machines between sessions. Its readiness and provisioning model does not provide automatic online expansion of one returning agent’s durable whole RootFS; persistent external task outputs have a separate scope.

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.

Task outputs and external connected storage can retain selected results, but reset pooled guest state is not a paused personal OS filesystem with arbitrary supported offline file operations.

See paused file operations →

Pool readiness does not mean session-state retention

The official security model describes dynamically assigned pooled machines that reset after every session. Agent identity and policy travel with the session; arbitrary previous guest state does not. Export required task outputs through the workflow before returning the Cloud PC to the pool.

Source 1 ↗

Consumption rounding differs from a second meter

The US Geography reference is $0.40 per consumption hour. Microsoft totals task runtime within each hourly window and rounds up to a full hour. Do not automatically interpret this as a separate one-hour charge per task or as billing solely for elapsed wall-clock time without the published aggregation rule.

Source 1 ↗

Always available adds a readiness charge

The US always-available option adds $5 per Cloud PC-month to consumption billing. It purchases immediate availability, not durable personal-session state. Billing is linked to Azure, with current geography, orchestration, identity and other product entitlements to verify for the selected integration.

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.

Windows 365 for Agents

Pricing reviewed 2026-10-19

US Geography consumption reference
$0.40 per hour
Task runtime is aggregated in hourly windows and rounded up to the next full hour under the published rule.
Rate source ↗
US always-available add-on
$5 per Cloud PC-month
Additional to consumption; pays for readiness, not previous-session disk/RAM retention.
Rate source ↗
Other geographies/integrations
Confirm current regional and product terms
Verify supported orchestration, licenses/entitlements and any other service charges.
Rate source ↗

These are Windows 365 for Agents references, not personal Windows 365 Business/Enterprise seat prices. Geography and integration affect eligibility and total cost. Do not count the monthly readiness fee as prepaid compute credit or promise durable session state from always-available wording.

Check current Windows 365 for Agents 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 Windows 365 for Agents

  1. Inventory agent integration, guest OS, identity/policies, pool behavior, task-output export and geography-specific billing rules.
  2. Move compatible Linux agent stages and durable files while retaining required native Windows or Microsoft-integrated task execution.
  3. Test repeated-state requirements, reset/export behavior, RootFS growth, paused files and combined pool consumption/readiness plus Linux runtime/storage costs.

Common questions

Is Sandbox0 a drop-in Windows 365 for Agents replacement?

Evaluate API compatibility and product scope. These products serve different core workflows, so a change may require a different architecture.

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 ↗learn.microsoft.com/en-us/windows-365/agents/identity-security ↗learn.microsoft.com/en-us/windows-365/agents/pricing-paygo-always-available ↗learn.microsoft.com/en-us/windows-365/agents/billing-w365a ↗learn.microsoft.com/en-us/windows-365/agents/introduction-windows-365-for-agents ↗Sandbox0 RootFS capacity →

Related comparisons

Bytebot vs Sandbox0 →CloudAxis vs Sandbox0 →Coasty vs Sandbox0 →ComputerUse.space vs Sandbox0 →Cyberdesk Windows Automation vs Sandbox0 →Google Agent Platform Sandbox (Code Execution / Shell / Computer Use) vs Sandbox0 →