boat vs Orgo
Orgo and boat both give each AI agent its own virtual machine with a desktop, a browser and a terminal. boat's sandboxes serve apps and webhooks on public HTTPS URLs, go up to 16 vCPU without a sales call, and run coding agents you steer by API. Orgo adds Windows, Mac and phone desktops, live-memory forks, and computers by the hour with no monthly plan.
Where boat stands out
Machines the internet can reach
Every boat sandbox serves up to 50 public HTTPS URLs, token-protected by default, and has a public IPv6 address. Orgo's docs say a computer has no public inbound hostname, so webhooks cannot reach it by default. Each Orgo computer gets direct SSH on a public host and port, which can change when the computer moves.
More vCPU per machine, no sales call
boat sandboxes go up to 16 vCPU / 32 GB on the $100 plan and above. Orgo stops at 6 vCPU / 32 GB self-serve and 8 vCPU / 64 GB on request.
Coding agents you drive from outside
Claude Code, Codex and five other agents are installed, and you prompt, steer and follow them through one API. With the optional AI credentials proxy on ($1 a day plus $1 per 1,000 requests), real keys of supported model providers stay 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
- Regions
- EU: Germany, Finland, France
Orgo
Orgo sells cloud computers for AI agents: each agent gets its own desktop that keeps working after you log off.
Every Orgo computer is a virtual machine with a full desktop, a browser and a terminal. You launch it from a dashboard or create it by REST, SDKs, a CLI or MCP. You can set up Hermes, OpenClaw, Claude Code or Codex in one click, or have Orgo's own computer-use agent drive the screen through an OpenAI-style chat API. You pay by the hour from credit with no plan, or buy a monthly plan that gives a vCPU and RAM pool shared by your computers.
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 workload | boat: price rank | Orgo: price rank |
|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month A coding agent that runs a dev server for others needs public inbound access. boat gives up to 50 HTTPS URLs per sandbox, token-protected by default. On Orgo you set it up: mark a template port public and serve TLS on the mapped host port, or forward a port over its direct SSH. | 5th cheapest of 365 $634 a month | 9th cheapest of 365 $1,355 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB An assistant that lives on its machine all month fits both. Orgo's OpenClaw guide says a computer has no public inbound hostname, so webhooks cannot reach it by default; you run the bot in outbound mode, or you set it up: a template port marked public maps to a host port, and you serve TLS on it. On boat every sandbox can take webhooks on a public HTTPS URL; you turn auto-stop off first (default 1 hour). | 6th cheapest of 361 $657 a month | 11th cheapest of 361 $1,179 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month | 189th of 336, with limits $2,000 a month | 156th of 336, with limits $579 a month |
| Sandboxes for your users 300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month Both cover 300 user sandboxes self-serve. Orgo runs computers past 50 by the hour from credit, up to 999 per account. boat's $100 plan allows 300 at once, with 30 starts a minute and 840 a day included. | 3rd cheapest of 349 $209 a month | 17th cheapest of 349 $665 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 | 60th of 349, with limits $3,134 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 | 37th cheapest of 359 $539 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month CI jobs that want more than Orgo's 6 vCPU self-serve (8 on request) fit boat sandboxes up to 16 vCPU / 32 GB. | 2nd cheapest of 374 $86 a month | 271st of 374, with limits $352 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB Both fit a box that runs all month. Orgo bills by the hour or with a monthly vCPU and RAM pool, and a stopped hourly computer pays a small disk rate. boat bills per second, has no compute charge for a stopped sandbox, and keeps its disk for at least 14 days (longer with a saved snapshot). | 19th cheapest of 371 $26 a month | 28th cheapest of 371 $40 a month |
| Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month Both have a desktop for computer use. Orgo adds Windows, Mac and phone screens and its own computer-use agent API. boat streams a Linux desktop at 1920x1080 and 60 fps. | 7th cheapest of 367 $200 a month | 14th cheapest of 367 $495 a month |
| Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month Both run a real Chrome or Chromium on a desktop. Orgo leaves from data-center IPs and can route browser traffic through your own HTTP CONNECT proxy. boat gives each sandbox a public IPv6 address and outbound IPv4; the IP changes at each resume, and boat has no built-in proxy. | 7th cheapest of 356 $200 a month | 15th cheapest of 356 $297 a month |
| Scraping & bots (startup scale) 50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month Both leave from data-center addresses. Orgo lets you route browser traffic through your own HTTP CONNECT proxy. boat has no dedicated IPv4 and no built-in proxy; the IP changes at each resume. | 6th cheapest of 343 $90 a month | 27th cheapest of 343 $268 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 | 161st of 327, with limits $5,363 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 | 7th cheapest of 351 $230 a month |
| GPU eval jobs 10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month | Not ranked: misses a requirement | Not 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 requirement | 20th of 35, with limits $534 a month |
Key features
| Feature | boat | Orgo |
|---|---|---|
| Isolation | Full VM with its own kernel, on physical hosts dedicated to boat (AMD Ryzen 7 9700X and Ryzen 9 9950X); vCPUs are shared | A virtual machine per agent with a full desktop |
| Biggest machine | Up to 16 vCPU / 32 GB / 251 GB on the $100 plan and above, no contact needed | Up to 6 vCPU / 32 GB / 150 GB self-serve (XL); 8 vCPU / 64 GB / 300 GB on request (2XL) |
| Machines at once (self-serve) | 100 to 2,000 sandboxes by plan | 50 computers on a paid plan or with bought credit; past 50 they run by the hour from credit, up to 999 per account |
| Start limits | Included starts per minute / hour / day: 12 / 60 / 200 on $20, up to 90 / 600 / 2,400 on $2,000; extra starts $0.03 to $0.07 each; above 6 times the included number a start fails | Creates per minute / hour / day: 10 / 100 / 500 on Starter, 30 / 300 / 2,000 on Growth, 60 / 1,000 / 5,000 on Scale |
| Reaching the machine from the internet | Up to 50 public HTTPS URLs per sandbox, token-protected by default; public IPv6; SSH over a shared IPv4 host and port | No public inbound hostname by default; you set it up: a template port marked public maps to a host port; direct SSH on a public host and port, which can change when the computer moves |
| Coding agents | Seven agents installed, with a prompt, steer and events API | One-click setup of Hermes, OpenClaw, Claude Code or Codex in a terminal session; its chat API drives the screen with Orgo's own agent |
| Keeping API keys out of the machine | Optional AI credentials proxy ($1 a day it is on plus $1 per 1,000 requests): real keys of supported model providers stay outside; other secrets still enter the sandbox | Vault secrets written into the computer at /root/.env |
| Desktop and computer use | Linux desktop and Chrome at 1920x1080, 60 fps, computer-use tools, embeddable stream | Full desktop over VNC you can embed with noVNC, plus Orgo's own computer-use agent API |
| Operating systems | Linux only; Windows programs through Wine | Linux, Windows, macOS, Android and physical iPhones |
| Files after a stop | Disk saved every minute and at each stop; no compute charge while stopped; disk kept at least 14 days, longer with a saved snapshot | Whole disk archived on stop and cold-booted on a new host; a stopped hourly computer pays a small disk rate ($0.0027/h for a Medium); plan computers can also sleep with memory kept |
| Fork | Fork a running sandbox from its last 1-minute snapshot; the copy boots fresh | Fork a running Linux plan computer with its live memory, or clone its disk |
| Outbound rules | Open internet, no provider-side allowlist | Per-computer domain, IP or CIDR allow and block rules in templates; route browser traffic through your own HTTP CONNECT proxy |
| GPUs | None; graphics run on the CPU | Small vGPU slices (2 or 4 GB VRAM) on Linux; price not published |
| Regions | Germany, Finland and France; you cannot pick; no US or Asia region | US region codes only (iad, sjc, lax); the region field is not used for placement; no EU region documented |
| Dedicated hosts and self-hosting | No self-hosting | Dedicated hosts by contract on Enterprise |
| Compliance and transparency | SOC 2 is in progress. We're open about where we stand: live status and data retention policy. | Not published |
Every feature Battleships tracks (43)
| Feature | boat | Orgo |
|---|---|---|
| Full VM (own kernel) | Yes | Yes |
| Docker inside | Yes | Yes |
| Nested virtualization | Yes | Not published |
| Root | Yes | Yes |
| Systemd | Yes | Not published |
| Desktop GUI | Yes | Yes |
| Browser | Yes | Yes |
| Computer-use API | Yes | Yes |
| SSH | Yes | Yes |
| Persistent disk | Yes | Yes |
| Pause/resume | Yes | Yes |
| Automatic snapshots | Yes | No documented way |
| Fork/clone | Yes | Yes |
| Live fork (no pause) | Yes | Yes |
| Memory fork | No documented way | Yes |
| Custom image (Docker or snapshot) | Yes | Yes |
| ≥24 h sessions | Yes | Yes |
| Idle auto-stop | No documented way | Yes |
| Wake on request | No documented way | Yes |
| HTTPS preview URLs | Yes | With your own setup |
| Raw TCP inbound | Yes | With your own setup |
| Public IPv4 | No documented way | No documented way |
| Egress allowlist | No documented way | Yes |
| Secret proxy | Yes | No documented way |
| Secret proxy for any API | Yes | No documented way |
| Preinstalled agents | Yes | Yes |
| Agent harness API | Yes | No documented way |
| MCP server | No documented way | Yes |
| GPU | No documented way | Yes |
| Windows | No documented way | Yes |
| Arm64 | No documented way | No documented way |
| Self-hosting | No documented way | Not published |
| BYOC | No documented way | Not published |
| Open source | No documented way | No documented way |
| Scoped API keys | Yes | Yes |
| Spend limits | Yes | Yes |
| Usage per sandbox (API) | Yes | No documented way |
| Audit logs | No documented way | Not published |
| SSO | No documented way | Not published |
| SOC 2 | No documented way | Not published |
| HIPAA | No documented way | Not published |
| EU data residency | No documented way | No documented way |
| Zero data retention | Yes | Not published |
Performance
Orgo 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 that run ran on an AMD Ryzen 9 9950X host. The boat fleet also has Ryzen 7 9700X hosts, which that 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. Orgo 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.
| Scenario | boat, a month | Orgo, 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 | $46 · Starter |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $26 · $20/mo | $40 · Plan pool (fee in the tier) |
| 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 | $900 · Starter |
| 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 | $347 · Growth |
| 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 | $946 · Scale · 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; unused time expires at the end of the billing period. A sandbox bills per second while it runs: $0.036/h 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 while sandboxes are stopped. Starts above the plan's included number cost $0.03 to $0.07 each.
Orgo: Orgo bills by the hour from credit, by the second with no hourly minimum: Medium (2 vCPU / 8 GB) $0.055/h, Large (4 vCPU / 16 GB) $0.11/h, XL (6 vCPU / 32 GB) $0.202/h while running, and a small disk rate while stopped. New accounts get $5 of credit. Monthly plans give a vCPU and RAM pool: Starter $29 (2 vCPU / 8 GB), Growth $99 (8 / 32 GB), Scale $399 (36 / 128 GB), 20% less when billed yearly. Plan resources cost the same whether they run or not. Add-ons cost $2.50 per 0.5 vCPU, $2.50 per GB of RAM and $1 per 20 GB of disk, each per month.
Limits in those scenarios
Orgo's limits are 50 computers on a paid plan, plus hourly computers up to 999 per account; 6 vCPU / 32 GB per computer self-serve (8 / 64 on request); creates per day of 500 on Starter, 2,000 on Growth and 5,000 on Scale; and agent runs at once of 5, 20 or 50 by plan. So 50 agents at 4 vCPU and 300 user sandboxes fit self-serve. 1,000 short jobs at once go past 999 computers and Scale's 5,000 creates a day, so they need Enterprise. boat's limits are sandboxes at once (100 to 2,000 by plan) and starts per minute, hour and day. The 1,000-job burst fits the $2,000 plan's 2,000 at once, but needs about 6,667 starts a day against 2,400 included. Extra starts cost $0.03 to $0.07 each, and starts above 270 a minute fail, so boat also misses that scenario as published.
| Scenario | boat | Orgo |
|---|---|---|
| One coding agent, workdays | None for this scenario |
|
| One agent that never stops | None for this scenario |
|
| 50 agents in parallel | None for this scenario |
|
| Sandboxes for your users | None for this scenario |
|
| 1,000 short jobs at once |
|
|
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 Orgo when
your agents must work on Windows, Mac or phone screens, need a live-memory fork or sleep, need outbound allowlists or small GPUs, must run in the US, or run only a few hours a month with no plan.
- Windows, Mac and phone desktops. Useful if your agents must work in Windows or Mac apps, or on an Android or iPhone screen.
- Fork and sleep with live memory. Useful for trying two next steps from the same moment: a fork keeps open apps, windows and running processes.
- Outbound rules, GPUs and dedicated hosts. Orgo can limit which hosts a computer reaches, attach small GPU slices, and run on dedicated hosts for Enterprise customers.
- By the hour with no plan. Useful for light or occasional use: no monthly fee and $5 of starting credit.
Use boat when
your agents build, test and serve software: Linux VMs up to 16 vCPU with public HTTPS URLs, and coding agents you drive by API, with an optional proxy that keeps model keys outside.
- Machines the internet can reach. A preview link, an OAuth callback or a Stripe webhook works on boat without extra setup. On Orgo you set it up: mark a template port public, serve TLS on the mapped host port, and update the address when it changes.
- More vCPU per machine, no sales call. A big build or test suite that scales with cores finishes sooner on more vCPU. A job that needs more than 32 GB of RAM fits Orgo's 2XL on request, not boat.
- Coding agents you drive from outside. Orgo writes your secrets into /root/.env inside the computer, where the agent can read them. Its chat API drives the screen with Orgo's own agent, so you script the coding agent's terminal yourself.
Questions people ask
Is boat cheaper than Orgo?
In four of the five scenarios we priced, yes. One coding agent at 4 vCPU on workdays costs $20 a month on boat and $46 on Orgo. One agent running all month costs $26 versus $40. 50 agents in parallel cost $317 versus $900, and 300 user sandboxes $200 versus $347. For 1,000 short jobs at once, Orgo costs $946 and boat $8,800, mostly paid extra starts, and neither fits that scenario as published. If 2 vCPU is enough, an Orgo Medium by the hour costs about $11 for the workday scenario with no plan, less than boat's $20 minimum.
Can I host a website or receive webhooks on an Orgo computer?
Yes, with your own setup. Orgo's docs say a computer has no public inbound hostname, so webhooks cannot reach it by default and bots should run in outbound mode. You can mark a template port public, which maps it to a host port, and serve TLS on it yourself. Every computer also has direct SSH on a public host and port. Every boat sandbox serves up to 50 public HTTPS URLs, token-protected by default, without extra setup.
Does boat run Windows or macOS like Orgo?
No. boat runs Linux VMs, with Wine for some Windows programs. Orgo offers Windows on Scale or with a $15/mo licence, and Macs ($100/mo) and iPhones ($65/mo) on paid plans while in stock.
Can I run Claude Code or Codex on both?
Yes. Orgo sets them up in one click and runs them in a terminal session on the computer. boat has seven coding agents installed and lets you prompt, steer and follow them through an API. With boat's optional, paid AI credentials proxy on, real keys of supported model providers stay outside the sandbox.
Where does my data run?
boat sandboxes run on machines in Germany, Finland and France, and you cannot pick one. Snapshot copies can be stored outside the EU. Orgo documents only US region codes (iad, sjc, lax), and its region field is not used for placement.
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