boat vs Vultr
A Vultr Cloud Compute instance is a plain KVM server that you set up, secure and keep running yourself. boat gives an agent a full VM too, with snapshots every minute, free stops, fork of a running machine, HTTPS ports, a desktop and coding agents already built in.
Where boat stands out
A disk saved every minute, and fork while running
boat saves a sandbox's disk every minute and when it stops, and a stopped sandbox costs nothing. You can fork a running sandbox in one call. A stopped Vultr instance is billed in full until you destroy it, and its snapshots are taken by hand and can take up to 30 minutes.
Coding agents and a desktop built in
Every boat sandbox has Claude Code, Codex, OpenCode and other agents ready, and you drive them from outside with a prompt, steer and events API. It also streams a 60fps desktop with Chromium for computer use. The credentials proxy keeps real AI keys out of the machine and can switch to another account when one runs out of credits.
HTTPS on up to 50 ports, no setup
Each boat sandbox can serve up to 50 ports on HTTPS URLs, protected by a token, plus raw TCP. A Vultr instance has a public IPv4 with any port open to it, but HTTPS means bringing a domain and running your own proxy and certificates.
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
Vultr
Vultr is a general-purpose cloud, and Cloud Compute is its line of KVM virtual servers.
A Vultr instance is a Linux or Windows VM you rent by size, with root, a public IPv4, local SSD or NVMe and a monthly transfer allowance. You choose shared vCPU plans (Regular, High Performance, High Frequency) or dedicated vCPU plans (Optimized Cloud Compute, VX1), and run your own software next to Vultr block storage, file systems, VPCs, Kubernetes, bare metal and GPUs. It is a server host, not an agent sandbox, so this page compares the two where they overlap: giving an agent its own Linux machine.
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 | Vultr: price rank |
|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month Both give a full KVM machine with root and Docker, but a Vultr instance starts empty and stays billed when stopped, while boat comes with agents, snapshots and free stops. | 3rd cheapest of 367 $634 a month | 114th of 367, with limits $1,060 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB A Vultr instance can run an assistant around the clock, but you build the agent setup, the HTTPS endpoint and the backups yourself; boat has them built in. | 5th cheapest of 366 $657 a month | 144th of 366, with limits $1,000 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Neither fits short bursts well: Vultr bills every started hour and each new server needs its own setup, and boat's start limits queue big bursts. | 186th of 345, with limits $2,000 a month | 168th of 345, with limits $1,000 a month |
| Sandboxes for your users 300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month Vultr's default instance limit is not published and is raised by request, so hundreds of user sandboxes need a ticket first; boat allows 100 to 2,000 by plan. | Cheapest of 361 $209 a month | 98th of 361, with limits $899 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 | 179th of 360, with limits $20,000 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 | 47th cheapest of 364 $893 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month Both run Docker builds on a full VM; on Vultr you also run the runner software, the image updates and the clean-up between jobs. | 4th cheapest of 377 $86 a month | 91st cheapest of 377 $857 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB Both fit a 24/7 dev box: a Vultr instance is a classic always-on server, and a boat machine keeps its disk with snapshots every minute. | 18th cheapest of 368 $26 a month | 28th cheapest of 368 $40 a month |
| Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month boat streams a 60fps desktop with Chromium and computer use; on a Vultr instance you install and run a desktop and VNC yourself. | 6th cheapest of 369 $200 a month | 252nd of 369, with limits $1,190 a month |
| Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month | 7th cheapest of 367 $200 a month | 300th of 367, with limits $894 a month |
| Scraping & bots (startup scale) 50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month | 6th cheapest of 346 $90 a month | 175th of 346, with limits $89 a month |
| Mass scraping & bots (undetectable) 1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month | 71st of 338, with limits $1,800 a month | 72nd of 338, with limits $1,818 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 | 168th of 360, with limits $150 a month |
| GPU eval jobs 10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month Only Vultr fits: it rents Nvidia and AMD GPUs, and boat has no GPUs. | 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 | Vultr |
|---|---|---|
| What you get | A sandbox API for agents, on full VMs | A general KVM server you set up and run yourself |
| Isolation | Every sandbox is a full VM with its own kernel | Every instance is a KVM VM with its own kernel |
| CPU | Dedicated AMD Ryzen 9 9950X (Zen 5) hosts; 26.69 runs/s at 4 vCPU / 8 GB | Shared vCPU plans, or dedicated AMD EPYC vCPU on Optimized and VX1 plans; no comparable measurement |
| Docker and nested virtualization | Docker, root, systemd and /dev/kvm on every machine, documented | Root and Docker; virtualization support is documented for VX1 and bare metal, not for the other Cloud Compute plans |
| Stopping | A stopped sandbox costs nothing and keeps its disk | A stopped instance is billed in full; you destroy it to stop paying |
| Snapshots | Automatic snapshots every minute and on stop | Manual disk snapshots that can take up to 30 minutes; paid scheduled backups |
| Fork | Fork a running sandbox in one call (disk, no memory) | Take a snapshot, then deploy new instances from it (disk, no memory) |
| Reaching the machine | HTTPS URLs for up to 50 ports, token-protected; raw TCP; own IP | Public IPv4 on any port; HTTPS needs your own domain, proxy and certificates |
| Desktop and computer use | 60fps desktop and Chromium, computer use and browser control, embeddable stream | Headless Linux; you install a desktop and VNC yourself |
| Coding agents | Claude Code, Codex, OpenCode and more preinstalled, with a prompt, steer and events API | None; you install and supervise agents yourself |
| Keeping keys out of the machine | AI credentials proxy for model providers and any HTTPS API you register; switch a credential to another account when credits run out | No credentials proxy; you put secrets on the server yourself |
| Outbound rules | Open internet on every plan; no provider-side allowlist | Vultr Firewall filters incoming traffic only; outbound rules go in the guest |
| GPUs and Windows | No GPUs; Linux guests only, Wine for Windows apps | Nvidia and AMD GPUs; Windows Server instances for a license fee |
| Regions | EU only: Germany, Finland, France | 33 locations on six continents |
| 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 | Vultr |
|---|---|---|
| Full VM (own kernel) | Yes | Not published |
| Docker inside | Yes | Yes |
| Nested virtualization | Yes | Not published |
| Root | Yes | Yes |
| Systemd | Yes | Yes |
| Desktop GUI | Yes | No |
| Browser | Yes | No |
| Computer-use API | Yes | No |
| SSH | Yes | Yes |
| Persistent disk | Yes | Yes |
| Pause/resume | Yes | No |
| Automatic snapshots | Yes | No |
| Fork/clone | Yes | Yes |
| Live fork (no pause) | Yes | Not published |
| Memory fork | No | No |
| Custom image (Docker or snapshot) | Yes | Yes |
| ≥24 h sessions | Yes | Yes |
| Idle auto-stop | No | No |
| Wake on request | No | No |
| HTTPS preview URLs | Yes | No |
| Raw TCP inbound | Yes | Yes |
| Public IPv4 | No | Yes |
| 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 | Not published |
| GPU | No | Yes |
| Windows | No | Yes |
| Arm64 | No | No |
| Self-hosting | No | No |
| BYOC | No | No |
| Open source | No | No |
| Scoped API keys | Yes | Not published |
| Spend limits | Yes | Not published |
| Usage per sandbox (API) | Yes | No |
| Audit logs | No | Yes |
| SSO | No | Yes |
| SOC 2 | No | Not published |
| HIPAA | No | Not published |
| EU data residency | Yes | Yes |
| Zero data retention | Yes | Not published |
Performance
There is no comparable measurement for Vultr Cloud Compute: it was not part of the public StarSling benchmark run where boat did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB. Vultr sells shared and dedicated vCPU plans on different Intel and AMD generations, so results would depend on which plan you pick.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Vultr 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 | Vultr, 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 | $10 · Usage account |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $26 · $20/mo | $40 · Usage account |
| 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 | $524 · Usage account |
| 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 | $894 · Usage account |
| 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 | $5,960 · Usage account |
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.
Vultr: Each Vultr plan has a fixed hourly price, billed by the started hour. Most plans are capped at 672 hours a month, so the monthly price is the most one instance can cost; VX1 plans have no cap and bill every hour used. Billing runs from deploy to destroy, including while the instance is stopped, and snapshots, backups and extra IPs cost extra. The prices here use the shared vCPU Regular plans.
Limits in those scenarios
Vultr does not publish how many instances a new account can run at once; you see your limit in the console and ask support to raise it, based on usage and billing history. So 50 agents in parallel, 300 user sandboxes and 1,000 short jobs at once all fit on price but likely need a ticket first. On boat the limit is a sandbox count (100 to 2,000 by plan) plus starts per minute, so the 1,000-job burst gets queued and needs more starts per day than the $2,000 plan allows.
| Scenario | boat | Vultr |
|---|---|---|
| One coding agent, workdays | None that bite | None that bite |
| One agent that never stops | None that bite | None that bite |
| 50 agents in parallel | None that bite | None that bite |
| Sandboxes for your users | None that bite | None that bite |
| 1,000 short jobs at once |
| None that bite |
Which one should you use?
Use Vultr when
you need GPUs, Windows, regions outside the EU, or a long-lived server that sits next to block storage, private networks and bare metal in one cloud.
- Regions around the world. Useful if your users are outside Europe: Vultr runs in 33 locations, from Seattle and Sao Paulo to Tokyo and Sydney.
- GPUs and Windows. Useful if your agents train or serve models, or need Windows: Vultr rents Nvidia and AMD GPUs and Windows Server instances.
- A whole cloud around the server. Useful if the agent's machine must sit next to block storage, shared file systems, private networks and bare metal in one account.
Use boat when
you want each agent on its own full VM that is ready to work: coding agents, a desktop, HTTPS ports and minute-by-minute snapshots built in, fork while running, and stops that cost nothing.
- A disk saved every minute, and fork while running. Agents sit idle for hours between tasks. On Vultr you either keep paying for an idle server or script snapshot, destroy and redeploy yourself, and lose whatever changed since your last snapshot.
- Coding agents and a desktop built in. A Vultr instance starts as an empty server, so you install, update and supervise each agent, desktop and API key yourself, and a key on the disk is a key the agent can leak.
- HTTPS on up to 50 ports, no setup. An agent that starts a dev server wants to share a working link right away. On Vultr that link is a bare IP with no TLS and no access token until you set them up on each machine.
Questions people ask
Is boat cheaper than Vultr?
It depends on how the machine runs. One coding agent on workdays comes to $10 a month on Vultr against $20 on boat, but one agent that never stops comes to $26 on boat against $40 on Vultr. 50 agents in parallel come to $317 on boat against $524 on Vultr, and sandboxes for your users come to $200 against $894.
Can I run Docker and nested VMs on Vultr?
Docker works, since every instance is a full KVM VM with root. Vultr documents virtualization support for its VX1 dedicated plans and bare metal, but not for the shared Regular, High Performance and High Frequency plans. On boat, Docker, systemd and /dev/kvm are documented on every machine.
Do I pay for a Vultr instance that is stopped?
Yes. Vultr keeps billing a stopped instance because its CPU, memory, disk and IP stay reserved, until you destroy it. On boat, a stopped sandbox costs nothing and keeps its disk.
Can I copy a running Vultr server?
You take a snapshot of its disk, which can take up to 30 minutes, then deploy new instances from it. On boat you fork a running sandbox in one call. Neither copies memory, so programs start fresh in the copy.
Where does my data run?
boat runs only in the EU (Germany, Finland and France), with EU data residency and a zero data retention option. Vultr has 33 locations, including Amsterdam, Paris, Frankfurt, London, Madrid, Milan, Stockholm and Warsaw in Europe.
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.
Start a sandbox
Install the CLI, then run boat new.
7-day free trial
More comparisons
- boat vs Oracle Cloud
- boat vs IONOS Cloud
- boat vs AWS AgentCore
- boat vs Contabo
- boat vs Azure Virtual Machines
- boat vs AWS Fargate
- boat vs netcup
- boat vs AWS EC2
- boat vs E2B
- boat vs Daytona
- boat vs Modal
- boat vs Vercel Sandbox
- AWS EC2 vs Vultr
- Google Compute Engine vs Vultr
- Azure Virtual Machines vs Vultr
- Oracle Cloud vs Vultr
- DigitalOcean vs Vultr
- Hetzner Cloud vs Vultr
- Vultr vs Linode
- Vultr vs Scaleway