boat vs Beam

Both give an agent a Linux machine you drive from code. Every boat sandbox is a full VM with /dev/kvm, a desktop and coding agents installed. Beam is a serverless GPU platform with fast gVisor sandboxes and a separate line of persistent CPU VMs that have a desktop.

Where boat stands out

A full VM with /dev/kvm in every sandbox

Each boat sandbox is a full VM with its own kernel, root, systemd and Docker. It also exposes /dev/kvm, so Android emulators and nested VMs run inside it. A Beam sandbox is a gVisor container by default. Beam VMs have their own kernel and systemd, but Beam documents no /dev/kvm for them.

Token-gated URLs on every sandbox

boat serves HTTPS on up to 50 ports per sandbox, token-gated by default, on every sandbox. A Beam sandbox gives each exposed port a public HTTPS URL with no built-in auth. Beam VMs add protected URLs and TCP ports on a persistent IPv6 address.

Agents installed, with one API

Claude Code, Codex and five other agents come installed, and you prompt, steer and follow them through one 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.

Beam

Beam is a serverless cloud for GPU inference, sandboxes, VMs and background jobs, built on its open-source runtime Beta9.

Beam runs functions, REST endpoints, task queues, pods, sandboxes and VMs on CPUs and GPUs, from a Python or TypeScript SDK. A sandbox is a container isolated with gVisor by default. It lives until you terminate it or its time to live runs out, and you can save it as a filesystem image or a memory snapshot. Beam VMs are a separate CPU product: a microVM with its own kernel, systemd, a root disk that stays when it stops, and a browser desktop. A VM can pause with its memory, fork and take disk snapshots. CPU work runs on Spot, Shared or Performance tiers at different prices.

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 rankBeam: price rank
Software factory
100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month
A software factory needs Docker, systemd and a disk that stays between runs. A boat sandbox and a Beam VM both give these. boat adds /dev/kvm for emulators and nested VMs, which Beam does not document.
5th cheapest of 365
$634 a month
7th cheapest of 365
$1,200 a month
Always-on AI employees
50 machines 24/7 · 2 vCPU / 4 GB
An assistant that lives on its machine all month fits a boat sandbox or a Beam VM. Both keep their disk while stopped. A Beam VM can also pause with its memory; boat adds coding agents installed and an agent API.
6th cheapest of 361
$657 a month
12th cheapest of 361
$1,495 a month
Code interpreter bursts
100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month
Beam fits this well: sandboxes bill per millisecond while they run and end on their own when the time to live runs out. boat bills per second, but each run counts as a start against the plan's start limits.
189th of 336, with limits
$2,000 a month
21st cheapest of 336
$161 a month
Sandboxes for your users
300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month
Both fit. boat gives each user a full VM with token-gated URLs and a saved disk. Beam gives each user a quick gVisor sandbox, with public URLs that your app must protect.
3rd cheapest of 349
$209 a month
14th cheapest of 349
$567 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
332nd of 349, with limits
$984 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
16th cheapest of 359
$258 a month
CI runners
40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month
CI jobs that build Android apps or VM images need /dev/kvm. A boat sandbox has it; Beam documents it on neither sandboxes nor VMs.
2nd cheapest of 374
$86 a month
29th cheapest of 374
$300 a month
24/7 dev boxes
1 machine 24/7 · 4 vCPU / 8 GB
Both keep your work when the machine stops. boat saves the disk every minute and at stop, and keeps it for at least 14 days. A Beam VM keeps its root disk (50 GiB by default, up to 100 GiB).
19th cheapest of 371
$26 a month
44th cheapest of 371
$55 a month
Computer-use agents
50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month
boat streams a desktop with Chrome at 1920x1080 and 60 fps in every sandbox, with computer tools for agents. Beam VMs also have a browser desktop with Chrome and an SDK for screenshots, clicks, typing and recording; Beam documents it for VMs, not sandboxes.
7th cheapest of 367
$200 a month
12th cheapest of 367
$431 a month
Browser-use agents
50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month
7th cheapest of 356
$200 a month
73rd of 356, with limits
$272 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
16th cheapest of 343
$196 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
44th of 327, with limits
$967 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
6th cheapest of 351
$219 a month
GPU eval jobs
10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month
Beam has GPUs, from RTX 4090 to H100 and B200. boat has no GPU on any size.
Not ranked: misses a requirementCheapest of 102
$547 a month
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

FeatureboatBeam
IsolationFull VM with its own kernel, on a physical host dedicated to boat (AMD Ryzen 7 9700X or Ryzen 9 9950X); vCPUs are sharedgVisor container by default for sandboxes, with an opt-in VM-backed mode; Beam VMs are CPU microVMs with their own kernel
Nested virtualizationYes, /dev/kvm in every sandbox for Android emulators and VMsNot documented for sandboxes or VMs
Docker and systemdDocker, systemd and root in every sandboxBeam VMs have Docker and systemd; sandboxes run Docker with docker_enabled, without systemd
LifetimeAuto-stop defaults to 1 hour; set up to 30 days, or turn it off on a paid plan for no session capSandbox default 10 minutes, or -1 to run until you terminate it; VMs run until you stop them, with optional idle sleep
Disk between runsDisk is saved every minute and at stop; a stopped sandbox keeps it at no charge for at least 14 days. Memory and running programs are not savedA sandbox keeps files through a saved image, a memory snapshot or a volume; a Beam VM keeps its root disk (50 GiB by default, up to 100 GiB) while stopped
Fork and snapshotsFork a running sandbox from its last 1-minute disk snapshot; memory is not keptSandbox memory snapshots with running processes; VMs pause with RAM kept, fork and take disk snapshots
GPUsNo GPU on any sizeGPU sandboxes, serverless GPUs and GPU machines up to H100, H200 and B200
Desktop and browserDesktop with Chrome in every sandbox, streamed at 1920x1080 and 60 fps, with computer tools for agentsBeam VMs have a browser Linux desktop (Chrome on x86, Firefox on ARM) and a Python/JS SDK for screenshots, clicks, typing and recording; documented for VMs, not sandboxes
Inbound accessUp to 50 HTTPS URLs per sandbox, token-gated by default; raw ports on public IPv6, and on IPv4 where the sandbox has its own address; the IP changes at each resumeSandboxes: a public HTTPS URL per port, no built-in auth. VMs: token-protected URLs and TCP ports on a persistent public IPv6 address; no inbound IPv4 listed
Outbound rulesOpen internet, no provider-side allowlistBlock all outbound traffic, or allow up to 64 hosts, wildcard domains, IPs or CIDRs, changeable while running
Coding agentsSeven agents installed, with a prompt, steer and events APIYou set it up: install agents in the sandbox or VM and run them through process exec; Beam's MCP server plugs into Claude Code, Codex and Cursor
Keeping API keys out of the sandboxOptional AI credentials proxy ($1 a day plus $1 per 1,000 requests): keys of supported model providers stay outside; other secrets still enter the sandboxBrokered secrets in sandboxes and VMs: the guest sees a placeholder, and Beam puts the real key into HTTPS headers to approved hosts
Price, 4 vCPU / 8 GB per hour$0.036, billed per second, from a $20 planPerformance $0.28 (default for new accounts), Shared $0.088, Spot $0.036 (interruptible with 30 s notice)
RegionsGermany, Finland and France; you cannot pick a region; no US or Asia regionUS by default; another region if you choose one; any region of your own cloud with BYOC
ComplianceSOC 2 in progress; no HIPAA BAA documentedSOC 2 Type II; HIPAA BAA for Enterprise customers
Self-hostingNoBeta9 runtime is AGPL-3.0 and free to self-host; bring your own cloud for a fee
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)
FeatureboatBeam
Full VM (own kernel)YesNo documented way
Docker insideYesYes
Nested virtualizationYesNo documented way
RootYesYes
SystemdYesNo documented way
Desktop GUIYesNo documented way
BrowserYesNo documented way
Computer-use APIYesNo documented way
SSHYesYes
Persistent diskYesWith your own setup
Pause/resumeYesNo documented way
Automatic snapshotsYesNo documented way
Fork/cloneYesYes
Live fork (no pause)YesYes
Memory forkNo documented wayYes
Custom image (Docker or snapshot)YesYes
≥24 h sessionsYesYes
Idle auto-stopNo documented wayNo documented way
Wake on requestNo documented wayNo documented way
HTTPS preview URLsYesYes
Raw TCP inboundYesNo documented way
Public IPv4No documented wayNo documented way
Egress allowlistNo documented wayYes
Secret proxyYesYes
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 wayYes
BYOCNo documented wayYes
Open sourceNo documented wayYes
Scoped API keysYesNot published
Spend limitsYesNot published
Usage per sandbox (API)YesNo documented way
Audit logsNo documented wayNot published
SSONo documented wayNo documented way
SOC 2No documented wayYes
HIPAANo documented wayYes
EU data residencyNo documented wayNot published
Zero data retentionYesNot published

Performance

boat28.06 runs/s
BeamNot measured

There is no comparable CPU measurement. Beam is not in StarSling's run of 3 October 2026 (run 37073680030), so we cannot say which is faster at CPU work. In that run 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 some sandboxes on AMD Ryzen 7 9700X hosts, which this run did not measure. For start time, ComputeSDK measured a Beam sandbox at 534 ms median with 100 at once (2026-11-06), at Beam's default size. boat is not in ComputeSDK.

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. Beam 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 monthBeam, 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$20 · Developer
One agent that never stops
4 vCPU / 8 GB, 50 GB disk, running 24/7 all month
$26 · $20/mo$58 · Developer
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$836 · Team
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$312 · Team
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$527 · Team · with limits

boat: Plans cost $20, $100, $500 or $2,000 a month. The whole plan price comes back as VM time for that month, and it does not carry over. A default sandbox (4 vCPU / 8 GB) costs $0.036 an hour, billed per second while it runs. A stopped sandbox has no compute charge, and boat keeps its disk for at least 14 days. Saved snapshots above the included count and the AI credentials proxy still charge while sandboxes are stopped. Create, fork and resume count as starts; the $20 plan includes 12 a minute and 200 a day, and extra starts cost $0.03 to $0.07 each.

Beam: Beam bills sandboxes per millisecond while they run, for each vCPU and GiB, on Spot, Shared or Performance tiers; new accounts default to Performance. For 4 vCPU / 8 GiB that is $0.28 an hour on Performance, $0.088 on Shared and $0.036 on Spot. VMs bill per second while running, and nothing for compute while stopped or paused. Developer is $20 a month with $20 of credit and 30 vCPUs at once; Team is $89 a month plus usage. There are no egress fees, and storage is not billed yet (planned $0.0001 per GB-hour).

Limits in those scenarios

Beam's Developer plan allows 30 vCPUs at once across sandboxes, VMs and apps, so 50 parallel agents of 4 vCPU need the Team plan. Beam does not bill storage yet, so the scenarios count no disk cost for it. Beam Spot can stop a machine with 30 seconds of notice, and a Spot VM loses its memory. boat's start limits are lower for bursts: on the $20 plan boat allows 12 starts a minute and 200 a day, and bills extra starts at $0.03 to $0.07 each. For 1,000 short jobs at once, neither fits as listed: boat's $2,000 plan charges for starts above 90 a minute and refuses them above 270, and Beam's Team plan allows 1,000 vCPUs at once where the burst needs 2,000.

ScenarioboatBeam
One coding agent, workdaysNone for this scenario
  • disk beyond 0 GiB: price unknown
One agent that never stopsNone for this scenario
  • disk beyond 0 GiB: price unknown
50 agents in parallelNone for this scenario
  • disk beyond 0 GiB: price unknown
Sandboxes for your usersNone for this scenario
  • disk beyond 0 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
  • Team: max 1,000 vCPU running at once (needs 2,000)
  • disk beyond 0 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 Beam when

your agents need GPUs, when you want serverless GPU endpoints, sandboxes and VMs in one platform, when a VM must pause with its memory, when you need hosting in the US or a SOC 2 report, or when you must self-host.

  • GPUs in the same platform. For workloads that need a GPU, Beam offers GPU sandboxes, serverless GPUs and GPU machines from RTX 4090 to H100, H200 and B200.
  • VMs that pause with their memory. A Beam VM can pause with its RAM and processes kept, fork, take disk snapshots and sleep when idle.
  • Outbound rules, compliance and US hosting. Teams that need a sandbox locked to a few hosts, a compliance report or US hosting get all three.
  • Open source and your own cloud. Its Beta9 runtime is AGPL-3.0 and free to self-host, and it can also run in your own cloud for a management fee.

Use boat when

your agents do real work over hours or days: installs, builds, Docker, emulators and a browser on a full VM with /dev/kvm, whose disk boat saves every minute, at a lower hourly price than Beam's tiers that are not interruptible.

  • A full VM with /dev/kvm in every sandbox. Agents install whatever a task needs. In a gVisor sandbox, tools that expect a real kernel or systemd can fail; on Beam you move that work to a VM. An Android emulator or a nested VM needs /dev/kvm, which Beam does not document on either product.
  • Token-gated URLs on every sandbox. A preview from a short-lived Beam sandbox is open to anyone with the URL unless the app adds its own auth, or you use a Beam VM. On boat the same preview is gated by default. A boat sandbox's IP changes at each resume, while a Beam VM keeps its IPv6 address.
  • Agents installed, with one API. On Beam you install the agents and build the prompt and event harness yourself.

Questions people ask

Is boat cheaper than Beam?

It depends on the scenario and on Beam's tier. In this page's scenarios, one coding agent on workdays costs $20 a month on each. One agent that never stops costs $26 on boat and $58 on Beam. 50 agents in parallel cost $317 against $836, and sandboxes for your users cost $200 against $312. Beam Spot costs $0.036 an hour for 4 vCPU / 8 GiB, the same as boat, but Beam can interrupt it with 30 seconds of notice. For 1,000 short jobs at once, neither fits as listed: boat charges for starts above 90 a minute and refuses them above 270, and Beam Team allows 1,000 vCPUs at once.

Is a Beam sandbox a VM?

Not by default. Beam's guide says its sandboxes are gVisor containers, and Beam has an opt-in VM-backed mode. Beam also sells separate CPU VMs with their own kernel, systemd and a desktop. Every boat sandbox is a full VM with its own kernel and /dev/kvm.

What happens to files when a Beam sandbox stops?

A Beam sandbox ends when you terminate it or its time to live runs out, which is 10 minutes by default in Python. To keep files, you save a filesystem image or memory snapshot, or attach a volume or durable disk. A Beam VM keeps its root disk while stopped and can pause with its memory. boat saves the disk every minute and at stop, and keeps it for at least 14 days; memory and running programs are not saved.

Does Beam have a desktop for computer use?

Yes, on Beam VMs. They have a browser Linux desktop with Chrome and a Python/JS SDK for screenshots, clicks, typing and recording. boat has a desktop with Chrome in every sandbox, streamed at 60 fps, with computer tools for agents.

Can boat run GPUs like Beam?

No. boat has no GPU on any size. Beam offers GPU sandboxes, serverless GPUs and GPU machines, so pick Beam for work that needs a GPU.

Can I run Docker in each?

Yes. Beam supports Docker inside a sandbox with docker_enabled and inside its VMs, and boat runs Docker in every sandbox. Beam VMs and boat sandboxes both have systemd; boat also has /dev/kvm for emulators and nested VMs.

Is boat SOC 2 compliant?

Not yet. SOC 2 is in progress, and Beam has SOC 2 Type II today. Live status is at status.ascii.dev and our data retention policy is at docs.boat.dev/data-retention.

Start a sandbox

Install the CLI, then run boat new.

7-day free trial

More comparisons