boat vs Azure Container Apps

Azure Container Apps runs containers that scale to zero, code interpreter sessions and, since September 2026, Container Apps Sandboxes: microVMs that suspend with their memory. boat gives each agent a full Linux VM with Docker, /dev/kvm, SSH, a desktop and coding agents installed.

Where boat stands out

Docker, /dev/kvm and SSH in every sandbox

A boat sandbox is a full VM with its own kernel, root, systemd, Docker and /dev/kvm for Android emulators and nested VMs. You can SSH in. Azure Container Apps Sandboxes also run Docker in the azure-dev image, but Azure documents neither /dev/kvm nor SSH. Regular container apps forbid privileged mode.

A desktop and a browser built in

Every boat sandbox has a Linux desktop and Chrome, streamed at 1920x1080 and 60 fps, with computer-use tools for agents. On Azure you build Chromium, Xvfb and noVNC into your own disk image and publish its port.

Coding agents with a prompt API

Seven coding-agent harnesses come installed, Claude Code and Codex among them. You prompt, steer and follow them through one API. Azure Sandboxes offer Claude Code and Copilot CLI images that you drive through the generic exec API.

The companies and products

boat

boat.dev runs sandboxes for AI agents: each one is a full Linux VM with its own kernel.

You start a machine with one command or one API call, then SSH in, run Docker, open a desktop, or let a built-in agent (Claude Code, Codex, OpenCode and others) work in it. The disk stays between sessions, and a stopped sandbox costs nothing.

Azure Container Apps

Azure Container Apps is Microsoft's serverless container platform: you deploy a container image and Azure runs, scales and routes traffic to it.

Container Apps has three compute types for agents. Apps and jobs run Linux containers that can scale to zero. Dynamic Sessions hand out Hyper-V isolated code sessions from prewarmed pools. Container Apps Sandboxes, generally available since September 2026, are microVMs that suspend with their memory, take snapshots, clone and mount persistent volumes. It suits teams already on Azure that want Azure networking, identity and billing.

Which use cases each one fits

For each use case, Battleships prices every sandbox provider on the same workload and sorts them from the cheapest monthly bill up. “5th cheapest of 365” means 4 of the other 364 providers run that workload, with every feature it needs, for less. “With limits” means the provider runs it but misses something, so it is listed after every full match. Open a use case for its whole ranking and what each provider misses.

Use case and its workloadboat: price rankAzure Container Apps: price rank
Software factory
100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month
Both run Docker builds and tests. Container Apps Sandboxes ship Docker in the azure-dev image; Azure does not document /dev/kvm, so Android emulators and nested VMs are not documented there. Regular container apps forbid privileged mode. boat has Docker and /dev/kvm in every sandbox.
5th cheapest of 365
$634 a month
94th of 365, with limits
$408 a month
Always-on AI employees
50 machines 24/7 · 2 vCPU / 4 GB
Both fit. Azure Sandboxes suspend when idle and resume with memory and disk, and Data Disk or Blob volumes keep files. boat keeps the disk (not memory) and saves it every minute; a stopped disk is kept for at least 14 days, longer with a saved snapshot.
6th cheapest of 361
$657 a month
197th of 361, with limits
$936 a month
Code interpreter bursts
100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month
Azure fits this well: code interpreter sessions come from prewarmed pools, sit behind Hyper-V and plug into LangChain, LlamaIndex and Semantic Kernel.
189th of 336, with limits
$2,000 a month
148th of 336, with limits
$349 a month
Sandboxes for your users
300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month
Both fit. Azure Sandboxes suspend idle user machines with their memory. boat gives each user a machine with a desktop, SSH and /dev/kvm that has no compute charge while stopped.
3rd cheapest of 349
$209 a month
232nd of 349, with limits
$912 a month
RL rollouts
1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month
39th of 349, with limits
$2,000 a month
157th of 349, with limits
$10,735 a month
Agent evals
250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month
10th cheapest of 359
$200 a month
73rd cheapest of 359
$1,226 a month
CI runners
40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month
Both run Docker: Azure Sandboxes in the azure-dev image, boat in every sandbox. Android builds that need an emulator need /dev/kvm, which boat documents and Azure does not.
2nd cheapest of 374
$86 a month
255th of 374, with limits
$183 a month
24/7 dev boxes
1 machine 24/7 · 4 vCPU / 8 GB
boat gives SSH and a disk that keeps installed tools. Azure Sandboxes keep the disk across a stop and give a shell through the aca CLI or the portal; Azure documents no SSH.
19th cheapest of 371
$26 a month
224th of 371, with limits
$13 a month
Computer-use agents
50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month
boat streams a built-in Linux desktop at 60 fps with computer-use tools. On Azure you set it up: build Xvfb, a VNC server and noVNC into your own image and publish its port. Azure has no official example and no computer-use API.
7th cheapest of 367
$200 a month
223rd of 367, with limits
$612 a month
Browser-use agents
50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month
boat ships Chrome on a desktop in every sandbox. On Azure you set it up: build Chromium or Playwright into your own OCI image and import it as a sandbox disk image.
7th cheapest of 356
$200 a month
315th of 356, with limits
$790 a month
Scraping & bots (startup scale)
50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month
6th cheapest of 343
$90 a month
280th of 343, with limits
$209 a month
Mass scraping & bots (undetectable)
1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month
73rd of 327, with limits
$1,800 a month
136th of 327, with limits
$4,021 a month
Parallel red teaming
200 at once · 2 vCPU / 4 GB · 5,000 sessions of 30 min a month
3rd cheapest of 351
$200 a month
207th of 351, with limits
$145 a month
GPU eval jobs
10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month
Not ranked: misses a requirementNot ranked: misses a requirement
macOS agents
5 at once · 4 vCPU / 8 GB · 1,000 sessions of 20 min a month
Not ranked: misses a requirementNot ranked: misses a requirement

Key features

FeatureboatAzure Container Apps
IsolationFull VM with its own kernel on a physical host dedicated to boat; vCPUs are sharedContainers for apps; Hyper-V per dynamic session; a microVM with its own kernel per sandbox
Docker and nested virtualizationDocker, systemd and root on every sandbox, plus /dev/kvm for Android emulators and VMsApps and jobs: no privileged mode, so no Docker daemon. Sandboxes: Docker in the azure-dev image, Buildah in ubuntu; /dev/kvm not documented
Files after a stopDisk saved every minute and at each stop; memory and running programs are not saved. A stopped disk is kept for at least 14 days, longer with a saved snapshotApps: files kept only in Azure Files mounts. Sandboxes: suspend with memory and disk, plus Data Disk or Blob volumes
ForkFork a running sandbox from its last 1-minute snapshot; disk onlySandboxes: snapshot a running sandbox with its memory and create new sandboxes from it; app replicas start fresh from the image
Machine sizeUp to 16 vCPU / 32 GB / 251 GB disk; the largest size needs the $100 plan or higherApps: up to 4 vCPU / 8 GiB on Consumption. Sandboxes: tiers up to 16 cores / 32 GiB / 320 GiB
Getting a shellSSH, plus CLI and API execInteractive shell through the aca CLI or the portal; exec API for sandboxes and sessions; no SSH documented
Desktop, browser and computer useBuilt-in Linux desktop and Chrome streamed at 1920x1080, 60 fps, with computer-use toolsYou set it up: build Chromium, Xvfb and noVNC into your own disk image and publish the port; no official example
Inbound accessUp to 50 HTTPS URLs per sandbox, token-gated by default. Public IPv6 on every sandbox; some sandboxes have no IPv4 of their own, so raw ports there are IPv6 onlyApps: HTTPS ingress with custom domains and IP rules; TCP ingress in a VNet environment. Sandboxes: published HTTPS ports and Private Endpoints
Idle handling and wakeAuto-stop timer, 1 hour by default, up to 30 days or off on paid plans. It does not watch for idle; a stopped sandbox is started by APIApps scale to zero and wake on the next request. Sandboxes suspend after an idle timeout (default 5 minutes) and wake on traffic
Start limits$20 plan: 12 starts a minute and 200 a day included; paid extra starts up to 1,200 a day, then refused. Create, fork and resume each countNo start limit or start fee documented; 300 active cores per region by default and 3,000 API requests a minute
Outbound rulesOpen internet, no provider-side allowlistSandboxes: egress proxy with allow and deny rules by domain or CIDR. Sessions: egress off by default. Apps: Azure Firewall and NSG rules in a VNet
GPUsNo GPU on any sizeApps: serverless NVIDIA T4 and A100, one per replica. Sandboxes: no GPU tier
Coding agentsSeven coding-agent harnesses installed, with an API to prompt and steer themSandboxes: Claude Code and GitHub Copilot CLI images, driven through the exec API. Sessions: LangChain, LlamaIndex and Semantic Kernel tools
Keeping API keys out of the sandboxOptional paid proxy ($1 a day plus $1 per 1,000 requests) keeps keys of supported AI providers out; a credential can switch accounts when credits run outSandboxes: egress Transform rules add a secret or managed-identity token to outbound headers for any HTTP host, one rule per host. Apps and sessions: secrets as env vars or files
RegionsGermany, Finland and France; no US or Asia region, and you cannot pick oneAbout 40 Azure regions on six continents
Compliance and transparencySOC 2 is in progress. We're open about where we stand: live status and data retention policy.SOC 2
Every feature Battleships tracks (43)
FeatureboatAzure Container Apps
Full VM (own kernel)YesNo documented way
Docker insideYesNo documented way
Nested virtualizationYesNo documented way
RootYesYes
SystemdYesNo documented way
Desktop GUIYesNo documented way
BrowserYesNo documented way
Computer-use APIYesNo documented way
SSHYesNo documented way
Persistent diskYesYes
Pause/resumeYesNo documented way
Automatic snapshotsYesNo documented way
Fork/cloneYesNo documented way
Live fork (no pause)YesNo documented way
Memory forkNo documented wayNo documented way
Custom image (Docker or snapshot)YesYes
≥24 h sessionsYesYes
Idle auto-stopNo documented wayYes
Wake on requestNo documented wayYes
HTTPS preview URLsYesYes
Raw TCP inboundYesYes
Public IPv4No documented wayNo documented way
Egress allowlistNo documented wayYes
Secret proxyYesNo documented way
Secret proxy for any APIYesNo documented way
Preinstalled agentsYesNo documented way
Agent harness APIYesNo documented way
MCP serverNo documented wayYes
GPUNo documented wayYes
WindowsNo documented wayNo documented way
Arm64No documented wayNo documented way
Self-hostingNo documented wayNo documented way
BYOCNo documented wayNo documented way
Open sourceNo documented wayNo documented way
Scoped API keysYesNot published
Spend limitsYesNo documented way
Usage per sandbox (API)YesYes
Audit logsNo documented wayYes
SSONo documented wayYes
SOC 2No documented wayYes
HIPAANo documented wayYes
EU data residencyNo documented wayYes
Zero data retentionYesNot published

Performance

boat28.06 runs/s
Azure Container AppsNot measured

Azure Container Apps and its Sandboxes are not in the public benchmarks we use, so we cannot say which is faster. In StarSling's run of 3 October 2026 (run 37073680030), boat did 28.06 runs per second on the V8 Web Tooling Benchmark at 4 vCPU / 8 GB, second of 15. That boat sandbox ran on an AMD Ryzen 9 9950X host. boat also places sandboxes on AMD Ryzen 7 9700X hosts and does not promise a CPU model.

V8 Web Tooling Benchmark at a 4 vCPU / 8 GB target, runs per second, higher is better. Source: StarSling's run of 3 October 2026 (run 37073680030), median of 5 sandboxes. Providers ran on different hardware in different regions. StarSling gives a shared rank when the run cannot separate two providers. Every boat sandbox in this run ran on an AMD Ryzen 9 9950X host. boat also places some sandboxes on AMD Ryzen 7 9700X hosts, which this run did not measure. Azure Container Apps is not in the public benchmarks we use, so we cannot say which is faster.

Pricing in five scenarios

Each bill is priced from the providers' public price lists by the open Battleships engine, with the cheapest plan that runs the scenario. Start limits, per-start fees and plan minimums are in the price, boat's included.

Scenarioboat, a monthAzure Container Apps, a month
One coding agent, workdays
4 vCPU / 8 GB, 50 GB disk, 8 hours a day, 22 days a month, files kept between days
$20 · $20/mo$0 · Pay-as-you-go · with limits
One agent that never stops
4 vCPU / 8 GB, 50 GB disk, running 24/7 all month
$26 · $20/mo$14 · Pay-as-you-go · with limits
50 agents in parallel
50 machines of 4 vCPU / 8 GB at once, 8-hour sessions, 1,100 sessions a month, Docker inside
$317 · $20/mo × 3 seats$259 · Pay-as-you-go · with limits
Sandboxes for your users
2 vCPU / 4 GB, 30,000 sessions of 10 minutes a month, up to 300 at once
$200 · $100/mo × 2 seats$912 · Pay-as-you-go · with limits
1,000 short jobs at once
2 vCPU / 4 GB, 1,000 at once, 3 minutes each, 200,000 jobs a month
$8,800 · $2000/mo · with limits$2,143 · Pay-as-you-go · with limits

boat: Plans cost $20, $100, $500 or $2,000 a month, and the whole plan price comes back as VM time for that month. You pay per second while a sandbox runs, busy or idle: $0.036 an hour for 4 vCPU / 8 GB. A stopped sandbox has no compute charge. Saved snapshots above the included count and the AI credentials proxy still charge. The $20 plan includes 12 starts a minute and 200 a day; extra starts cost $0.03 to $0.07, and starts above 6 times the included number fail.

Azure Container Apps: Consumption apps bill per second for the vCPU and memory you allocate while a replica runs, with a monthly free grant and nothing at zero replicas. Code interpreter sessions bill each allocated session in one-hour steps. Sandboxes bill per second on the same Consumption meters for their tier while running, $0.432 an hour for XL (4 cores / 8 GiB) in eastus, and pay no CPU or memory while stopped. Snapshot and volume storage will be billed at Premium Blob ZRS rates; Azure announces this as coming soon.

Limits in those scenarios

The price table on this page shows Container Apps apps and code interpreter sessions. They do not fit these scenarios: regular apps forbid Docker and keep files only in Azure Files, and sessions are short code runs. Container Apps Sandboxes fit, and our boat vs Azure Container Apps Sandboxes page prices them. At eastus list price, Sandboxes cost $68 a month for the workday agent (boat $20), $3,421 for 50 agents (boat $317) and $989 for sandboxes for your users (boat $200). An always-on XL sandbox costs about $315 a month, or $262 with a 3-year savings plan that Azure does not confirm for Sandboxes (boat $26). For 1,000 short jobs at once, Sandboxes cost $2,148 but need a quota increase above the default 300 active cores per region. boat does not fit that scenario: its $2,000 plan charges for starts above 90 a minute and refuses them above 270. Azure storage charges are not in these totals.

ScenarioboatAzure Container Apps
One coding agent, workdaysNone for this scenario
  • machine size not published
  • different kind of product: code interpreter
  • files survive a stop only in Azure Files mounts: packages installed elsewhere reset
  • machine size per session not published
One agent that never stopsNone for this scenario
  • machine size not published
  • different kind of product: code interpreter
  • files survive a stop only in Azure Files mounts: packages installed elsewhere reset
  • machine size per session not published
50 agents in parallelNone for this scenario
  • missing Docker inside
  • machine size not published
  • different kind of product: code interpreter
  • files survive a stop only in Azure Files mounts: packages installed elsewhere reset
Sandboxes for your usersNone for this scenario
  • machine size not published
  • different kind of product: code interpreter
  • machine size per session not published
  • disk beyond 4 GiB: price unknown
1,000 short jobs at once
  • 2,400 starts/day included, paid above that (needs ~6,667)
  • 600 starts/hour included, paid above that (needs ~1,000)
  • starts above 90/min are paid, and starts above 270/min fail with 429
  • heavy, sustained CPU use can be slowed by other tenants
  • disk beyond 4 GiB: price unknown
boat's own limits
  • Plans start at $20 a month. The plan price comes back as VM time for that month. Unused time does not carry over. Source
  • Starts (create, fork, resume): the $20 plan includes 12 a minute, 60 an hour and 200 a day. The $2,000 plan includes 90 a minute, 600 an hour and 2,400 a day. Extra starts cost $0.03 to $0.07. A start fails with 429 above 3 times the included starts per minute, or 6 times per hour or per day. Source
  • A stopped sandbox has no compute charge. boat keeps its disk for at least 14 days. A saved snapshot keeps it longer; extra saved snapshots and the AI credentials proxy are paid. Source
  • Auto-stop defaults to 1 hour. You can turn it off on a paid plan. The trial allows 2 hours at most. Source
  • boat saves the disk every minute and at each stop. Memory, running programs and /tmp are not saved: a resume is like a reboot. A machine failure loses writes after the last good snapshot. Source
  • Machines run in Germany, Finland and France only. Snapshot copies can be stored outside the EU. Source
  • Public IPv6 and outbound IPv4. No dedicated IPv4. The IP changes at each resume. Source
  • Hosts are AMD Ryzen 7 9700X and Ryzen 9 9950X. vCPUs are shared. No GPU. You cannot boot your own Docker image. Source

Which one should you use?

Use Azure Container Apps when

you already run on Azure and need serverless containers that scale to zero, GPUs for inference, short code interpreter calls behind Hyper-V, stateful microVM sandboxes that suspend with their memory, or regions outside the EU.

  • Stateful microVM sandboxes. For sandboxes that suspend with their memory, resume in under a second and clone from snapshots inside your Azure account.
  • Serverless GPUs. For inference or GPU jobs next to your agents: T4 and A100 GPUs billed per second that scale to zero.
  • Scale to zero and wake on traffic. Idle apps and sandboxes stop billing compute and wake on the next request; code sessions come from prewarmed pools.
  • Inside Azure, in about 40 regions. Teams already on Azure get Entra ID, VNets, private endpoints, egress rules and Azure billing, in about 40 regions on six continents.

Use boat when

your agents need a full VM with Docker, /dev/kvm for emulators, SSH and a built-in desktop and browser, with coding agents installed, at $0.036 an hour for 4 vCPU / 8 GB.

  • Docker, /dev/kvm and SSH in every sandbox. Agents that start an Android emulator or a nested VM need /dev/kvm. On Azure, that work moves outside Container Apps to a VM that you size and patch yourself.
  • A desktop and a browser built in. Computer-use and browser agents work on boat from the first start. On Azure you build, test and update the desktop image yourself.
  • Coding agents with a prompt API. You run agent tasks from outside the sandbox without writing your own wrapper around a shell.

Questions people ask

Is boat cheaper than Azure Container Apps?

For most agent workloads, yes, at list price. Against Azure Container Apps Sandboxes, boat costs $20 against $68 for a workday agent and $317 against $3,421 for 50 agents. For 1,000 short jobs at once, Azure Sandboxes fit at $2,148 and boat does not, because its plans cap starts. Below about $20 of use a month, Azure's no-fee pay-as-you-go costs less.

Does Azure Container Apps have agent sandboxes?

Yes. Container Apps Sandboxes became generally available in September 2026. Each sandbox is a microVM with its own kernel that suspends with its memory, takes snapshots and mounts volumes, in tiers up to 16 cores / 32 GiB.

Can I run Docker inside Azure Container Apps?

Not in a regular container app, which forbids privileged mode. Container Apps Sandboxes include Docker in the azure-dev image. Azure does not document /dev/kvm for emulators or nested VMs. Every boat sandbox runs Docker and exposes /dev/kvm.

How long can an Azure session or sandbox last?

A code interpreter session is deleted once it has been idle for its cooldown, between 300 and 3,600 seconds, and each code run stops after 220 seconds. Container Apps Sandboxes document no runtime cap. boat has no session cap once auto-stop is off (default 1 hour).

Do files survive a restart on Azure Container Apps?

In a regular container app, only files in an Azure Files mount survive. Container Apps Sandboxes keep memory and disk across a suspend, and Data Disk or Blob volumes keep files. boat saves the disk every minute and at each stop; memory and running programs are not saved.

Does Azure Container Apps have GPUs?

Yes, for apps. Serverless NVIDIA T4 and A100 GPUs can attach to a replica in some regions, billed per second. Sandboxes have no GPU tier. boat has no GPU on any size, so GPU work belongs on Azure or another GPU cloud.

Is boat SOC 2 compliant?

SOC 2 is in progress, and boat has no SOC 2 report today. We're open about where we stand: live status at status.ascii.dev and our data retention policy at docs.boat.dev/data-retention.

Start a sandbox

Install the CLI, then run boat new.

7-day free trial

More comparisons