boat vs exe.dev

boat and exe.dev both run each machine as a real VM with root, Docker and SSH. boat adds a built-in desktop, an automatic disk snapshot every minute and an API to drive the agents inside. exe.dev adds nine regions, boot from your own Docker image, an outbound lockdown and team features like SSO.

Where boat stands out

A built-in screen for your agents

Every boat sandbox can show a 60 fps desktop with Chromium. Agents click, type and browse through the computer-use and browser-control APIs. You can embed the live stream in your own app. exe.dev has no built-in desktop: Shelley's browser tool drives a headless Chromium, and for a full screen you install a desktop and VNC yourself.

State saved every minute, with nothing to set up

boat snapshots a running sandbox's disk every minute and again when it stops. If a machine fails, you lose what was written after the last good snapshot. Memory and running programs are not saved, so a resume is like a reboot. You can fork a running sandbox and keep saved snapshots as templates. exe.dev can back up a VM disk to your own S3 bucket and restore it (preview), but you trigger each backup yourself.

Drive agents inside from outside

boat ships seven agent harnesses and one API to prompt them, steer them mid-task and stream their events, with SDKs for TypeScript and Python. exe.dev preinstalls Claude Code, Codex and Pi too. Its own Shelley agent takes prompts through the API, and the HTTPS exec API runs the others and streams their output.

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.

exe.dev

exe.dev, made by Bold Software, Inc., rents Linux VMs for developers and AI agents that you reach over SSH and HTTPS.

You reserve a pool of vCPUs and memory, and your VMs share it. You can also start standalone VMs billed by the second. Each VM is KVM-isolated with Cloud Hypervisor, boots from a container image in about two seconds, and has root, systemd and Docker. It comes with Claude Code, Codex, Pi and exe.dev's own Shelley agent. It suits developers who want a persistent dev box or a sandbox for their agents.

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 rankexe.dev: price rank
Software factory
100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month
boat fits a software factory with a built-in desktop and automatic snapshots every minute. exe.dev runs the code too, but you install a desktop and trigger disk backups yourself.
5th cheapest of 365
$634 a month
267th of 365, with limits
$3,721 a month
Always-on AI employees
50 machines 24/7 · 2 vCPU / 4 GB
Both fit always-on assistants: each gets its own VM with a disk that lasts. exe.dev VMs run until you stop them. On boat, you turn auto-stop off (default 1 hour) on a paid plan.
6th cheapest of 361
$657 a month
20th cheapest of 361
$3,833 a month
Code interpreter bursts
100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month
Neither is built as a code interpreter, and both rank low for many tiny runs.
189th of 336, with limits
$2,000 a month
231st of 336, with limits
$175 a month
Sandboxes for your users
300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month
For sandboxes you hand to your own users, boat gives token-protected URLs and 100 to 2,000 sandboxes at once by plan, with daily start limits. exe.dev Personal allows 50 VMs in total; a Work pool allows up to 1,000 VMs, and pools above 32 vCPU need support approval.
3rd cheapest of 349
$209 a month
252nd of 349, with limits
$1,234 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
260th of 349, with limits
$5,250 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
280th of 359, with limits
$1,050 a month
CI runners
40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month
Both bill CI per second: boat sandboxes while they run, exe.dev standalone VMs while they run. exe.dev pools bill whether jobs run or not. boat's $20 plan includes 200 starts a day; extra starts cost $0.03 to $0.07 each.
2nd cheapest of 374
$86 a month
289th of 374, with limits
$504 a month
24/7 dev boxes
1 machine 24/7 · 4 vCPU / 8 GB
Both make good 24/7 dev boxes with root, SSH and a persistent disk. exe.dev offers nine regions and your own Docker image. boat offers a desktop in the browser.
19th cheapest of 371
$26 a month
24th cheapest of 371
$35 a month
Computer-use agents
50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month
Computer-use agents need a screen. boat streams a built-in 60 fps desktop. On exe.dev you set it up: install a desktop and VNC on the VM yourself.
7th cheapest of 367
$200 a month
270th of 367, with limits
$1,055 a month
Browser-use agents
50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month
boat gives agents a visible Chromium and a browser-control API. On exe.dev, Shelley's built-in tool drives a headless Chromium, and other agents need a browser you set up.
7th cheapest of 356
$200 a month
187th of 356, with limits
$530 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
187th of 343, with limits
$278 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
165th of 327, with limits
$5,490 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
51st of 351, with limits
$264 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

Featureboatexe.dev
IsolationFull VM with its own kernel, on physical hosts dedicated to boat (AMD Ryzen 7 9700X and Ryzen 9 9950X); vCPUs are sharedKVM VM with Cloud Hypervisor on rented bare metal; you cannot choose the kernel
Desktop and computer useBuilt-in 60 fps desktop and Chromium, computer use, browser control, embeddable streamNo built-in desktop or stream; you can install a desktop and VNC yourself, and Shelley's browser tool takes screenshots of a headless Chromium
Saved stateDisk snapshot every minute while running and at each stop; memory and running programs are not saved; a stopped sandbox's disk is kept at least 14 days, a saved snapshot has no scheduled expiryDisk persists and is replicated across the regional cluster; on-demand incremental disk backups to your own S3 bucket, with restore to a new VM (preview); no scheduled backups or user rollback points
Fork and templatesFork a running sandbox from its last 1-minute snapshot; saved snapshots as templates, shareable by codecp copies a VM's disk and config into a new VM
Custom imagesStart from boat's image or a saved snapshot; you cannot boot your own Docker or OCI imageBoot from your own Docker image
Preinstalled agentsClaude Code, Codex, pi, OpenCode, Prime Agent, Kimi Code, Mistral VibeClaude Code, Codex, Pi and Shelley
Agent APIPrompt, steer and stream typed events for any built-in agent, with TypeScript and Python SDKsSend a prompt to Shelley from the CLI or HTTPS API; run other agents through the HTTPS exec API and read the streamed output
MCP serverNoYes, at exe.dev/mcp
Inbound trafficUp to 50 HTTPS URLs per sandbox, token-gated by default; SSH; a public IPv6 address with raw ports; no dedicated IPv4HTTPS proxy for ports 3000 to 9999, one of them public; SSH; no public IP or raw TCP
Outbound controlOpen internet; no egress allowlistOpen by default; an egress policy can block everything but attached integrations
SecretsOptional AI credentials proxy, off by default: $1 per day it is on plus $1 per 1,000 requests; covers model providers and HTTPS APIs you registerIntegrations inject secrets at the network edge for any HTTP host, S3, GitHub and a large catalog
Nested virtualizationYes, /dev/kvm on every sandboxOff by default; available on request from support
RegionsGermany, Finland and France; no US or Asia region; you cannot pick a region9 regions in the US, Europe, Asia and Australia
Teams and accessOrg seats, scoped API keys, spend limits, usage per sandbox; no SSOUnlimited users, SSO and roles on Work; BYOC and dedicated hardware on Enterprise
Session lengthAuto-stop defaults to 1 hour; set up to 30 days, or turn it off on a paid plan for no capNo cap; VMs run until you stop them
Compliance and transparencySOC 2 is in progress. We're open about where we stand: live status and data retention policy.No SOC 2 report published
Every feature Battleships tracks (43)
Featureboatexe.dev
Full VM (own kernel)YesYes
Docker insideYesYes
Nested virtualizationYesWith your own setup
RootYesYes
SystemdYesYes
Desktop GUIYesNo documented way
BrowserYesYes
Computer-use APIYesNo documented way
SSHYesYes
Persistent diskYesYes
Pause/resumeYesYes
Automatic snapshotsYesNo documented way
Fork/cloneYesYes
Live fork (no pause)YesYes
Memory forkNo documented wayNo documented way
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 wayWith your own setup
Secret proxyYesYes
Secret proxy for any APIYesYes
Preinstalled agentsYesYes
Agent harness APIYesYes
MCP serverNo documented wayYes
GPUNo documented wayNo documented way
WindowsNo documented wayNo documented way
Arm64No documented wayNo documented way
Self-hostingNo documented wayNo documented way
BYOCNo documented wayYes
Open sourceNo documented wayNo documented way
Scoped API keysYesYes
Spend limitsYesNot published
Usage per sandbox (API)YesYes
Audit logsNo documented wayNot published
SSONo documented wayYes
SOC 2No documented wayNo documented way
HIPAANo documented wayNo documented way
EU data residencyNo documented wayYes
Zero data retentionYesNot published

Performance

boat28.06 runs/s
exe.devNot measured

exe.dev is 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/s on the V8 Web Tooling Benchmark 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.

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. exe.dev 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 monthexe.dev, 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$35 · Personal
One agent that never stops
4 vCPU / 8 GB, 50 GB disk, running 24/7 all month
$26 · $20/mo$35 · Personal
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$1,848 · Personal
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$530 · Personal · 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$1,050 · Personal · 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 unused time expires at the end of the period. Sandboxes bill per second while they run, busy or idle: $0.036 an hour for 4 vCPU / 8 GB. A card is required, also for the 7-day trial.

boat: A stopped sandbox has no compute charge, and boat keeps its disk for at least 14 days. Saved snapshots above the plan's included count and the AI credentials proxy still charge while sandboxes are stopped. The $20 plan includes 200 starts a day; extra starts cost $0.03 to $0.07 each.

exe.dev: Pools are reserved capacity that your VMs share. Personal is a flat fee paid upfront, $15 to $155 a month for 2 to 16 vCPU. Work is $75 a month for the team ($37.50 for the first 3 months), plus $0.105 an hour per extra 2 vCPU / 4 GiB while it exists. A pool bills the same whether its VMs are busy or idle.

exe.dev: Standalone VMs bill per second, with no minimum, at $0.105 per 2 vCPU / 4 GiB per hour. A stopped standalone VM pays nothing for CPU and memory but keeps its disk and still counts toward your VM limit.

Limits in those scenarios

On exe.dev Personal you get up to 16 vCPU and 50 VMs in total, standalone VMs included, so the 300-at-once and 1,000-at-once scenarios don't fit there. On Work, a pool holds up to 1,000 VMs, pools above 32 vCPU need approval from exe.dev support, and standalone VMs stay off until a team admin turns them on. On boat, sandboxes at once go from 100 to 2,000 by plan, and starts per minute, hour and day are capped by plan. boat never queues a start: above the included starts each extra start costs $0.03 to $0.07, and above 6 times the included number a start fails with 429. So even the $2,000 plan does not fit 1,000 short jobs at once.

Scenarioboatexe.dev
One coding agent, workdaysNone for this scenario
  • flat pool: the tier is billed whether used or not
One agent that never stopsNone for this scenario
  • flat pool: the tier is billed whether used or not
50 agents in parallelNone for this scenarioNone for this scenario
Sandboxes for your usersNone for this scenario
  • Personal: max 50 concurrent
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
  • Personal: max 50 concurrent
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 exe.dev when

your users are outside Europe, or your team wants VMs booted from its own Docker image, an outbound lockdown, SSO and roles.

  • Regions around the world. Useful if your users are far from Europe: exe.dev runs in nine regions across the US, Europe, Asia and Australia.
  • Your own Docker image. exe.dev can boot each VM straight from your own Docker image.
  • Team and enterprise controls. Its Work plan covers a whole team with SSO and roles, and it serves an MCP server for controlling VMs.
  • Outbound lockdown. exe.dev can block a VM's internet access except for the integrations you attach.

Use boat when

your agents do real work for hours: builds, tests, Docker, a built-in desktop and a browser, with the disk saved every minute and one API to drive the agents inside.

  • A built-in screen for your agents. Computer-use agents work by looking at a screen. On exe.dev you set up and stream a desktop yourself before an agent can check a UI or click through a web app.
  • State saved every minute, with nothing to set up. Agents sometimes break things, like deleting files or wrecking an install. exe.dev can back up a VM disk to your own S3 bucket and restore it (preview), but you trigger each backup yourself. boat saves every minute with nothing to set up.
  • Drive agents inside from outside. On exe.dev only Shelley takes prompts through the API. You can run Claude Code or Codex through the HTTPS exec API and read its streamed output, but you cannot steer a task mid-run or get typed agent events.

Questions people ask

Is boat cheaper than exe.dev?

In our scenario table boat's monthly bill is lower in four of the five workloads, and the gap is largest for 50 agents running in parallel ($317 vs $1,848). For 1,000 short jobs at once exe.dev's bill is lower, but neither provider fits that load on the plans we priced: exe.dev Personal caps at 50 VMs, and boat's start limits refuse part of the burst.

Does exe.dev have a desktop or computer use?

Not built in. exe.dev VMs are reached over SSH and an HTTPS proxy. You can install a desktop and VNC yourself, and Shelley's browser tool drives a headless Chromium and takes screenshots. boat gives every sandbox a built-in 60 fps desktop with Chromium and APIs for computer use and browser control.

Does exe.dev have snapshots?

Partly. exe.dev has on-demand disk backups to your own S3 bucket with restore, in preview, and cp copies a VM. boat snapshots running sandboxes every minute with no setup and keeps saved snapshots as templates. On boat, memory and running programs are not saved.

Can I run Claude Code on both?

Yes. exe.dev preinstalls Claude Code, Codex and Pi along with its own Shelley agent. boat preinstalls Claude Code, Codex, pi, OpenCode, Prime Agent, Kimi Code and Mistral Vibe. With boat's paid credentials proxy on, the real keys of supported providers stay out of the sandbox.

Where do boat and exe.dev run?

boat sandboxes run on machines in Germany, Finland and France, and you cannot pick a region. Snapshot copies can be stored outside the EU. exe.dev has nine regions in the US, Europe, Asia and Australia.

Is boat SOC 2 compliant?

SOC 2 is in progress; there is no SOC 2 report 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