boat vs Steel

Steel is browser infrastructure: your agent asks its API for a cloud Chrome, drives it over CDP and releases it when the work is done. boat gives your agent a whole Linux VM with root, Docker, a disk that persists, a 60fps desktop with Chromium and coding agents built in.

Where boat stands out

A whole machine, not only a browser

A boat sandbox is a full Linux VM with SSH, root, systemd, Docker and /dev/kvm, 4 vCPU / 8 GB by default and up to 16 vCPU / 32 GB. A Steel session is a Chrome you reach through CDP and an API, with no shell and no published machine size. The agent's own code has to run somewhere else.

No session cap, and a disk that stays

On boat's paid plans a sandbox runs as long as you want, its disk persists with snapshots every minute and on stop, and a stopped sandbox costs nothing. You can fork a running sandbox to try two paths at once. A Steel session ends after 15 minutes on Launch, 1 hour on Scale and 24 hours on Enterprise.

Coding agents built in, driven from outside

boat ships Claude Code, Codex, OpenCode and other agents inside each sandbox, and you drive them from outside with a prompt, steer and events API. Steel's CLI, skills and MCP server give a coding agent a browser, but the agent itself runs on your machine. boat's credentials proxy also keeps real API keys 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.

Isolation: Full VM per sandbox, own kernel
Hardware: Dedicated AMD Ryzen 9 9950X (Zen 5) hosts
Regions: EU: Germany, Finland, France
Plans: From $20 a month, all of it usage credit, 7-day trial

Steel

Steel is an API that runs fleets of cloud browsers for AI agents: you create a session, connect Puppeteer, Playwright or Selenium, and release it when done.

Each Steel session is a headful Chromium in a VM in Steel's cloud, driven over CDP or through a computer-use input API, with a live viewer you can embed. Around it Steel sells managed proxies, CAPTCHA solving, saved browser profiles, a credentials vault that signs in for the agent, a files API and readable traces. It is a different kind of product from boat: a managed browser, not a general sandbox, so this page compares where the two overlap, which is agents that use a browser.

Kind of product: Browser infrastructure for AI agents: cloud browser sessions over an API
Isolation: Docs describe each session as a headful Chromium in a VM; the VM type is not named
Session length: Up to 15 minutes on Launch, 1 hour on Scale, 24 hours on Enterprise
Concurrency: 10 browsers on Launch, 100 on Scale, 1,000+ on Enterprise
Region: Browsers run in one region, us-east; proxies change only the visible IP location
Billing: Browser hours billed by the minute, rounded up: $0.10 an hour on Launch, $0.08 on Scale
Compliance: Scale and Enterprise include SSO and a HIPAA-ready BAA
Source code: The steel-browser runtime is open source under Apache-2.0, with about 7.8K GitHub stars

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. “3rd cheapest of 367” means 2 of the other 366 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 workload	boat: price rank	Steel: price rank
Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month Steel is a compromise here: a session gives you a browser, not a shell, so installs, builds, Docker and tests need a machine like boat's.	3rd cheapest of 367 $634 a month	124th of 367, with limits $1,115 a month
Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB An assistant that works all month runs into Steel's session cap of 1 hour on Scale, so it must start new sessions and reload its profile again and again.	5th cheapest of 366 $657 a month	238th of 366, with limits $3,070 a month
Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month	186th of 345, with limits $2,000 a month	264th of 345, with limits $333 a month
Sandboxes for your users 300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month 300 sandboxes at once is more than Steel's Scale limit of 100 browsers, and each user gets a browser rather than a machine.	Cheapest of 361 $209 a month	244th of 361, with limits $990 a month
RL rollouts 1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month	34th of 360, with limits $2,000 a month	257th of 360, with limits $4,150 a month
Agent evals 250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month	10th cheapest of 364 $200 a month	348th of 364, with limits $400 a month
CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month CI needs a shell, Docker and a known CPU size; Steel sessions offer none of these, so it is a compromise.	4th cheapest of 377 $86 a month	239th of 377, with limits $120 a month
24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB	18th cheapest of 368 $26 a month	241st of 368, with limits $73 a month
Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month Steel's computer-use API drives the browser only; boat streams a 60fps desktop where the agent can use any Linux app as well as Chromium.	6th cheapest of 369 $200 a month	148th of 369, with limits $367 a month
Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month Both fit: Steel hands your agent a cloud Chrome with proxies and CAPTCHA solving, and boat gives it Chromium on a full VM where it can also run code.	7th cheapest of 367 $200 a month	25th cheapest of 367 $470 a month
Scraping & bots (startup scale) 50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month Both fit at startup scale; Steel adds managed proxies and CAPTCHA solving, while boat runs the scraper, its queue and its database on the same machine.	6th cheapest of 346 $90 a month	56th cheapest of 346 $540 a month
Mass scraping & bots (undetectable) 1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month Steel fits mass scraping better: residential proxies, CAPTCHA solving and a stealth browser on Enterprise, while boat uses datacenter IPs and solves no CAPTCHAs.	71st of 338, with limits $1,800 a month	9th of 338, with limits $7,950 a month
Parallel red teaming 200 at once · 2 vCPU / 4 GB · 5,000 sessions of 30 min a month	2nd cheapest of 360 $200 a month	220th of 360, with limits $242 a month
GPU eval jobs 10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month	Can't run it	Can't run it
macOS agents 5 at once · 4 vCPU / 8 GB · 1,000 sessions of 20 min a month	Can't run it	Can't run it

Key features

Feature	boat	Steel
Kind of product	A full Linux VM per sandbox, with a desktop and Chromium inside	A managed Chromium browser per session, driven over CDP
Isolation	Every sandbox is a full VM with its own kernel	Each session is a Chromium in a VM, per the docs; VM type not named
Machine size	4 vCPU / 8 GB by default, up to 16 vCPU / 32 GB by plan	Not published; billed per browser hour
Shell, Docker and nested virtualization	SSH, root, systemd, Docker and /dev/kvm on every machine	No shell; you reach the browser through CDP and the REST API
Session length	No session cap on paid plans	15 minutes on Launch, 1 hour on Scale, up to 24 hours on Enterprise
Files after a session	The disk persists, with snapshots every minute and on stop; a stopped sandbox costs nothing	Files API keeps session files in org storage; profiles keep cookies and logins (300 MB, deleted after 30 days unused)
Fork	Fork a running sandbox into a copy	No fork; a new session can start from a saved profile or session context
Browser and computer use	60fps desktop and Chromium, computer use and browser control, embeddable stream	Headful Chrome, a computer-use input API, an embeddable live view with human takeover, mobile mode and traces
Bot detection	Datacenter IPs and Chromium; no CAPTCHA solving	Residential proxies, CAPTCHA solving and dedicated IPs; a stealth browser on Enterprise
Reaching the machine	HTTPS URLs for up to 50 ports, token-protected; raw TCP; dedicated IPv6 or IPv4	No inbound ports; a CDP websocket and an unauthenticated live view URL
Secrets	AI credentials proxy: real API keys stay outside the sandbox, for model providers and any HTTPS API you register	Credentials API (beta) stores web logins and signs in for the agent without showing it the password
Coding agents	Claude Code, Codex, OpenCode and more preinstalled, with a prompt, steer and events API	A CLI, skills and an MCP server give your coding agent a Steel browser; the agent runs elsewhere
Self-hosting	Managed cloud only	Open-source steel-browser (Apache-2.0) in Docker, and Steel Outpost inside your own network
Regions	EU only: Germany, Finland, France	us-east only; proxy location sets the visible IP
Compliance and transparency	SOC 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)

Feature	boat	Steel
Full VM (own kernel)	Yes	Yes
Docker inside	Yes	No
Nested virtualization	Yes	No
Root	Yes	No
Systemd	Yes	No
Desktop GUI	Yes	No
Browser	Yes	Yes
Computer-use API	Yes	Yes
SSH	Yes	No
Persistent disk	Yes	No
Pause/resume	Yes	No
Automatic snapshots	Yes	Yes
Fork/clone	Yes	No
Live fork (no pause)	Yes	No
Memory fork	No	No
Custom image (Docker or snapshot)	Yes	No
≥24 h sessions	Yes	Yes
Idle auto-stop	No	Yes
Wake on request	No	No
HTTPS preview URLs	Yes	No
Raw TCP inbound	Yes	No
Public IPv4	No	No
Egress allowlist	No	No
Secret proxy	Yes	No
Secret proxy for any API	Yes	No
Preinstalled agents	Yes	No
Agent harness API	Yes	No
MCP server	No	Yes
GPU	No	No
Windows	No	No
Arm64	No	No
Self-hosting	No	Yes
BYOC	No	Yes
Open source	No	Yes
Scoped API keys	Yes	Not published
Spend limits	Yes	Not published
Usage per sandbox (API)	Yes	No
Audit logs	No	Not published
SSO	No	Yes
SOC 2	No	No
HIPAA	No	Yes
EU data residency	Yes	No
Zero data retention	Yes	Not published

Performance

boat26.69 runs/s

SteelNot measured

There is no comparable measurement. Steel was not in StarSling's public CPU benchmark and does not publish how many vCPUs a browser session gets, so there is no number to set against boat's 26.69 runs per second at 4 vCPU / 8 GB. Steel says a session starts in under a second.

Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Steel has no published run in this suite yet, so there is no like-for-like number to compare.

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.

Scenario	boat, a month	Steel, 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	$250 · Scale · with limits
One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month	$26 · $20/mo	$250 · Scale · 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	$784 · Scale · 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	$510 · Scale · with limits
1,000 short jobs at once 2 vCPU / 4 GB, 1,000 at once, 3 minutes each, 200,000 jobs a month	$2,000 · $2000/mo · with limits	$950 · Scale · with limits

boat: You pick a monthly plan from $20, and all of it becomes usage credit spent per second on running sandboxes. A stopped sandbox costs nothing and keeps its disk.

Steel: Steel bills browser hours by the minute, rounded up, from create until release or timeout: $0.10 an hour on the free Launch plan and $0.08 on Scale, which costs $250 a month and includes $100 of usage each month. Proxy bandwidth, CAPTCHA solves and browser tools are billed on top.

Limits in those scenarios

The Steel prices in the table are for one browser session, the only thing it sells, so none of them is a 4 vCPU / 8 GB machine with a shell, Docker or a machine disk that keeps installed state. The scenarios that run for hours also need the Scale plan at $250 a month, and even then each session stops after 1 hour; only Enterprise goes to 24 hours. Scale caps you at 100 browsers at once, so 300 sandboxes for your users and 1,000 short jobs at once need Enterprise. On boat, the 1,000-job burst is the one that does not fit: bursts above 90 starts a minute are queued, and it needs more starts per day than the top plan allows.

Scenario	boat	Steel
One coding agent, workdays	None that bite	Scale: sessions capped at 1 h missing persistent disk machine size not published browser-session product, not a general sandbox
One agent that never stops	None that bite	Scale: sessions capped at 1 h missing persistent disk machine size not published browser-session product, not a general sandbox
50 agents in parallel	None that bite	Scale: sessions capped at 1 h missing persistent disk, Docker inside machine size not published browser-session product, not a general sandbox
Sandboxes for your users	None that bite	Scale: max 100 concurrent machine size not published browser-session product, not a general sandbox different kind of product: browser session
1,000 short jobs at once	max 2,400 starts/day (needs ~6,667) max 600 starts/hour (needs ~1,000) bursts above 90 starts/min are queued (a burst of 333 takes ~4 min to start)	Scale: max 100 concurrent machine size not published browser-session product, not a general sandbox different kind of product: browser session

Which one should you use?

Use Steel when

your agent only needs a browser: scraping behind proxies and CAPTCHAs, saved logins, human takeover, browsers in the US, or an open-source browser you host yourself.

Proxies, CAPTCHAs and stealth. Useful for scraping sites that block bots: Steel has residential proxies, CAPTCHA solving, dedicated IPs and a stealth browser on Enterprise.
Open source and self-hosted. Useful if browsers must run on your own machines: steel-browser is Apache-2.0 open source, and Steel Outpost runs inside your network.
Ready-made browser tooling, in the US. Useful if you want saved logins, a credential vault, human takeover and traces as ready parts, with browsers in the US.

Use boat when

your agent browses and also works: installs, builds, Docker and dev servers on a full VM with no session cap, a disk that stays, a 60fps desktop and coding agents built in.

A whole machine, not only a browser. An agent that browses and then installs packages, runs a dev server or builds in Docker needs a second machine next to its Steel browser, and you wire the two together and pay for both.
No session cap, and a disk that stays. A long agent run on Steel's Scale plan must be cut into one-hour sessions, each a new browser seeded from a saved profile or session context. Running work and anything in memory start over at every cut.
Coding agents built in, driven from outside. With Steel you still host the agent, its files and its API keys yourself, and keep that host alive. On boat the agent, its tools and its browser live on one machine you start with one call.

Questions people ask

Is boat cheaper than Steel?

The two price different things: the table sets a 4 vCPU / 8 GB boat VM against one Steel browser session with no shell or machine disk. One coding agent on workdays comes to $20 a month on boat and $250 on Steel, because sessions over 15 minutes need the Scale plan, and 50 agents in parallel come to $317 against $784. For 1,000 short jobs at once Steel comes to $950 against $2,000 on boat, and neither fits.

Can a Steel session run Docker or a coding agent?

No. A Steel session is a cloud Chrome you control over CDP and an API; it has no shell, so there is nowhere to run Docker or an agent's code. Steel's CLI and MCP server let a coding agent on your own machine use a Steel browser. On boat each sandbox is a full VM with Docker and coding agents already installed.

How long can a Steel session run?

Up to 15 minutes on Launch, 1 hour on Scale and 24 hours on Enterprise, and the timeout cannot be extended on a live session. Longer jobs start a new session and reload cookies and logins from a profile. boat has no session cap on paid plans.

Does boat have proxies or CAPTCHA solving like Steel?

No. boat runs Chromium on datacenter IPs and does not solve CAPTCHAs. If you scrape sites that block bots, Steel's residential proxies and CAPTCHA solver are built for that.

Can I use boat and Steel together?

Yes. An agent on a boat sandbox can open a Steel browser through its API and keep its code, files and tools on the boat VM. That puts proxied browsing on Steel and the rest of the work on a machine that keeps its disk.

Is boat SOC 2 compliant?

SOC 2 is in progress. We're open about where we stand: live status at status.ascii.dev and our data retention policy at docs.boat.dev/data-retention. Steel lists a HIPAA-ready BAA and SSO on its Scale and Enterprise plans.

Start a sandbox

Install the CLI, then run boat new.

7-day free trial