boat vs Cerebrium

These are different kinds of product. boat gives each agent its own full Linux VM to work inside. Cerebrium is a serverless GPU cloud that runs your model or app as containers. The containers scale up with requests and down to zero.

Where boat stands out

A whole machine to work inside

Each boat sandbox is a full VM with its own kernel, root, SSH, systemd and Docker. It also exposes /dev/kvm, so Android emulators and VMs run inside it. Cerebrium runs one app container that you reach through its endpoints or the Run API. It documents no shell to log into.

The disk stays, and you can fork it

boat saves the disk every minute and once more when the sandbox stops. A stopped sandbox has no compute charge, and boat keeps its disk for at least 14 days; a saved snapshot keeps it longer. You can fork a running sandbox from its last 1-minute snapshot. On Cerebrium, files persist in the /persistent-storage volume, and there is no fork.

Agents and desktop built in

Claude Code, Codex and five other agent harnesses are installed, and you prompt, steer and follow them through one API. Each sandbox has a 1920x1080, 60 fps desktop with Chrome for computer use. An optional paid credentials proxy keeps the real keys of supported AI providers outside the sandbox.

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.

Cerebrium

Cerebrium is a serverless GPU cloud for real-time AI: voice agents, video models, LLMs and other ML apps.

You describe an app in a cerebrium.toml file, with your Python code or your own Dockerfile. Cerebrium builds it and serves it as HTTP or WebSocket endpoints. The Cerebrium Run API also runs an uploaded Python file or function without an endpoint. Cerebrium starts containers when requests arrive, scales them across GPUs and regions, and scales them to zero when traffic stops. It is built for teams that serve models in production, not for agents that need a machine of their own. This page compares the places where the two overlap.

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 rankCerebrium: price rank
Software factory
100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month
A software factory needs Docker, root and a disk that is still there tomorrow. On Cerebrium you set it up yourself: bake packages into your Dockerfile and keep files in the shared /persistent-storage volume. Docker inside an app is not documented.
5th cheapest of 365
$634 a month
205th of 365, with limits
$2,008 a month
Always-on AI employees
50 machines 24/7 · 2 vCPU / 4 GB
An assistant that lives on one machine all month keeps its disk on boat, saved every minute. On Cerebrium you keep its state in /persistent-storage and its packages in the image. Both bill every hour the machine runs.
6th cheapest of 361
$657 a month
261st of 361, with limits
$2,534 a month
Code interpreter bursts
100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month
Short code runs fit Cerebrium's model: a container starts on a request, bills per second, and scales back to zero when the work stops. Cerebrium Run also executes an uploaded Python file or function on chosen CPU or GPU compute without an endpoint.
189th of 336, with limits
$2,000 a month
222nd of 336, with limits
$131 a month
Sandboxes for your users
300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month
Both can serve your users. boat gives each user a full VM with a saved disk and public URLs. Cerebrium gives them an endpoint that scales with traffic and down to zero.
3rd cheapest of 349
$209 a month
226th of 349, with limits
$821 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
78th of 349, with limits
$4,068 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
45th cheapest of 359
$639 a month
CI runners
40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month
CI jobs that build containers or Android apps need Docker and /dev/kvm. Every boat sandbox has both. Cerebrium does not document either inside an app.
2nd cheapest of 374
$86 a month
88th cheapest of 374
$766 a month
24/7 dev boxes
1 machine 24/7 · 4 vCPU / 8 GB
A dev box needs SSH and a disk that keeps your installs. Cerebrium documents no SSH, and files outside /persistent-storage reset when a container stops.
19th cheapest of 371
$26 a month
272nd of 371, with limits
$102 a month
Computer-use agents
50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month
boat ships a Linux desktop and Chrome streamed at 1920x1080, 60 fps, with computer tools for agents. Cerebrium has no desktop or computer-use product. You can install a headless browser in your own image, but there is no display stream or computer-use API.
7th cheapest of 367
$200 a month
217th of 367, with limits
$575 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
109th of 356, with limits
$572 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
105th cheapest of 343
$387 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
134th of 327, with limits
$3,991 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
234th of 351, with limits
$191 a month
GPU eval jobs
10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month
Cerebrium has GPUs from T4 to B200, up to 8 in a container. boat has no GPU on any size.
Not ranked: misses a requirement28th cheapest of 102
$1,975 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

FeatureboatCerebrium
What you getA full VM with its own kernel, on a physical host dedicated to boat (AMD Ryzen 7 9700X or Ryzen 9 9950X; vCPUs are shared). You or an agent work inside it. You cannot boot your own Docker imageAn app you deploy as containers built from your code or Dockerfile, served as endpoints
Shell accessSSH, root via sudo and systemd, like a normal Ubuntu 24.04 serverNo SSH documented; you reach the app through its HTTP or WebSocket endpoints
Docker and nested virtualizationDocker with BuildKit and Compose inside, and /dev/kvm for Android emulators and VMsThe app is the container; Docker inside and /dev/kvm are not documented
Files after a stopThe disk is saved every minute and at each stop. Memory, running programs and /tmp are not saved. A stopped sandbox has no compute charge, and boat keeps its disk for at least 14 daysFiles in the shared /persistent-storage volume stay (100 GB free, then $0.05 per GB a month). You bake packages into the image. Other changes reset
Fork and snapshotsFork a running sandbox from its last 1-minute snapshot; writes from the last minute or so can be missingNo fork. An early-beta memory checkpoint (CPU and GPU memory, no files) speeds up new containers of the same build
Run timeAuto-stop defaults to 1 hour. Set up to 30 days, or turn it off on a paid plan for no session capNo container cap, but a request times out at response_grace_period (15 minutes by default, 12 hours at most for async runs)
One-off code runsRun commands in a sandbox over SSH, the CLI or the API; it bills per second while it runsCerebrium Run uploads a tar of code and runs a Python file or function on CPU or a chosen GPU, without an endpoint
Scale to zero and wakeNo wake on request; you start a stopped sandbox by API. Auto-stop is a timer and does not watch for idleApps scale to zero and start again on the next request, with autoscaling up to your replica limit
GPUsNo GPU on any size; graphics run on the CPUT4 to B200 and RTX PRO 6000, up to 8 per container, plus Inferentia2 and Trainium
Inbound accessUp to 50 public HTTPS URLs per sandbox, token-gated by default. Public IPv6 on every sandbox; some sandboxes have no IPv4 of their own, so a raw port there is IPv6 onlyOne endpoint per app on one port, open unless you set disable_auth = false; custom domains on Standard
Desktop and browserLinux desktop and Chrome streamed at 1920x1080, 60 fps, with computer tools for agentsNo desktop or computer-use product. You can install a headless browser in your own image; there is no display stream or computer-use API
Coding agentsSeven agent harnesses installed, with a prompt, steer and events APINone to run inside; you install one in your own image. An Agent Skill and a docs MCP server help your own agent deploy to Cerebrium
Keeping API keys out of the machineOptional AI credentials proxy ($1 a day it is on, plus $1 per 1,000 requests). With it on, real keys of supported providers stay outside the sandbox. Other secrets still enter itSecrets are exposed to the app as environment variables
ComplianceSOC 2 in progress; no SOC 2 report today, no HIPAA BAA documentedSOC 2, HIPAA, GDPR and ISO 27001 listed
RegionsGermany, Finland and France; no US or Asia region, and you cannot pick oneUS and EU generally available; Asia, South America, Canada and the Middle East on request
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)
FeatureboatCerebrium
Full VM (own kernel)YesNo documented way
Docker insideYesNot published
Nested virtualizationYesNo documented way
RootYesNot published
SystemdYesNot published
Desktop GUIYesNo documented way
BrowserYesWith your own setup
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 inboundYesNot published
Public IPv4No documented wayNo documented way
Egress allowlistNo documented wayNo documented way
Secret proxyYesNo documented way
Secret proxy for any APIYesNo documented way
Preinstalled agentsYesNo documented way
Agent harness APIYesNo documented way
MCP serverNo documented wayNo documented way
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 limitsYesNot published
Usage per sandbox (API)YesYes
Audit logsNo documented wayYes
SSONo documented wayNo documented way
SOC 2No documented wayYes
HIPAANo documented wayYes
EU data residencyNo documented wayYes
Zero data retentionYesNot published

Performance

boat28.06 runs/s
CerebriumNot measured

Cerebrium is not in the public benchmarks we use, so we cannot say which is faster at CPU work. On the V8 Web Tooling Benchmark in StarSling's run of 3 October 2026 (run 37073680030), boat did 28.06 runs/s at 4 vCPU / 8 GB, second of 15. 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. Cerebrium's own speed claims are about GPU inference and cold starts, which this test does not cover.

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. Cerebrium 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 monthCerebrium, 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$53 · Hobby · with limits
One agent that never stops
4 vCPU / 8 GB, 50 GB disk, running 24/7 all month
$26 · $20/mo$210 · Hobby · 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$2,557 · Hobby · 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$715 · Hobby
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$1,681 · Standard

boat: Plans cost $20, $100, $500 or $2,000 a month, and a card is required. The whole plan price comes back as VM time for that month; it does not carry over. A 4 vCPU / 8 GB sandbox costs $0.036 an hour, billed per second from provisioning while it runs, busy or idle. The $20 plan includes 12 starts a minute and 200 a day; extra starts cost $0.03 to $0.07 each. A stopped sandbox has no compute charge. Saved snapshots above the included count and the AI credentials proxy still charge while sandboxes are stopped.

Cerebrium: Cerebrium bills per second for each vCPU, GB of memory and GPU while a container runs: $0.00000655 per vCPU-second and $0.00000222 per GB-second on the default tier. It does not bill the platform's cold-start time. The default tier can be interrupted; the protected tier costs 2x on all compute. Hobby is free plus compute. Standard is $100 a month on top of compute, with no usage credit. Storage past the first 100 GB costs $0.05 per GB a month.

Limits in those scenarios

In the coding agent, always-on and 50-agent scenarios, Cerebrium needs your own setup: packages go into the image and files into /persistent-storage, because other changes reset at a stop. The 50-agent fleet also needs Docker inside, which Cerebrium does not document. Cerebrium does not publish an egress price, so the scenarios count none; boat does not bill egress. boat's limits: no GPU, no US or Asia region, a fixed disk per size, and you lose writes after the last good snapshot if a machine fails. boat's start limits are lower for bursts. At 1,000 short jobs at once, the $2,000 plan includes 90 starts a minute, charges $0.03 to $0.07 for extra starts, and refuses starts above 270 a minute with 429. boat never queues a start. Cerebrium's Standard plan fits that burst.

ScenarioboatCerebrium
One coding agent, workdaysNone for this scenario
  • files survive a stop only in /persistent-storage: packages installed elsewhere reset
  • egress price not published
One agent that never stopsNone for this scenario
  • files survive a stop only in /persistent-storage: packages installed elsewhere reset
  • egress price not published
50 agents in parallelNone for this scenario
  • files survive a stop only in /persistent-storage: packages installed elsewhere reset
  • egress price not published
Sandboxes for your usersNone for this scenario
  • egress price not published
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
  • egress price not published
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 Cerebrium when

you need GPUs to serve a model, voice or video app as an endpoint that scales with traffic, when it must run in the US or Asia, or when you need a SOC 2 or HIPAA report today.

  • GPUs for serving models. For model inference, Cerebrium offers GPUs from T4 to B200, up to 8 in one container, billed per second.
  • Endpoints that scale to zero. Apps that wait for traffic can scale to zero, start on the next request, and grow to many replicas under load.
  • More regions and compliance reports. It runs in the US and the EU by default, with Asia, South America, Canada and the Middle East on request, and lists compliance reports.

Use boat when

your agents need a machine of their own: a shell, root, Docker, emulators and a desktop, on a full VM whose disk is saved every minute and comes back at the next start.

  • A whole machine to work inside. A coding agent installs tools, starts services and runs docker build as it goes. On Cerebrium you would wrap every step in an endpoint or Run call. Docker inside an app and /dev/kvm are not documented.
  • The disk stays, and you can fork it. An agent that pip installs, clones repos and edits config anywhere on the disk keeps that work on boat. On Cerebrium you set it up yourself: install packages in the Dockerfile and write work to /persistent-storage. Everything else resets when the container scales down.
  • Agents and desktop built in. On Cerebrium you build the agent image and the browser setup yourself, and there is no display stream. Its secrets arrive as environment variables inside the container. On boat, environment variables and secrets other than proxied AI keys also enter the sandbox.

Questions people ask

Is Cerebrium a sandbox for AI agents?

Not in the usual sense. Cerebrium is a serverless GPU cloud that runs your app as autoscaling containers behind endpoints, plus one-off runs through Cerebrium Run. It documents no SSH or desktop and ships no coding agent; you would install one in your own image. A boat sandbox is a full VM an agent works inside.

Is boat cheaper than Cerebrium?

In four of the five scenarios on this page, yes. For one 4 vCPU / 8 GB agent that never stops, boat costs $26 a month. At Cerebrium's list rates, the same size costs about $116 a month on the default interruptible tier and about $231 on the protected tier. For 1,000 short jobs at once, Cerebrium's Standard plan fits the burst, while boat's $2,000 plan charges for starts above 90 a minute and refuses them above 270.

Can boat run GPUs like Cerebrium?

No. boat has no GPU on any size. If your agent needs a GPU to serve or run a model, Cerebrium offers GPUs from T4 to B200.

What happens to files when a Cerebrium container stops?

Files in the /persistent-storage volume stay; the rest of the container resets. To keep packages, install them in your Dockerfile. boat saves the disk every minute and at each stop, and keeps it for at least 14 days while the sandbox is stopped. Memory and running programs are not saved.

Can I run Docker inside Cerebrium?

Cerebrium builds your app from a Dockerfile, but running Docker inside the app is not documented. A boat sandbox runs Docker, systemd and /dev/kvm like a normal Ubuntu server.

Is boat SOC 2 compliant?

SOC 2 is in progress, so there is no SOC 2 report today. No HIPAA BAA is documented. 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