boat vs Dedalus Labs
Both give each agent its own persistent Linux VM with root, systemd and its own kernel. Dedalus Machines sleep when idle and wake on the next command, while boat adds what an agent needs around the machine: public ports, snapshots and forks, a desktop, and coding agents you drive from outside.
Where boat stands out
Agents can host what they build
Every boat sandbox can serve HTTPS on up to 50 ports, token-protected by default, plus raw TCP and its own IP. Dedalus machines get private IPs and take no inbound connections, so the only way in is the API or an SSH port forward.
Snapshots, forks and custom images
boat saves the disk every minute and on every stop, lets you roll back to any snapshot, and forks a running sandbox into a copy. Dedalus keeps the root disk across sleep but has no snapshot, fork or image API.
Agents, a desktop and keys built in
Claude Code, Codex and five other agents come preinstalled, and you prompt, steer and follow them from outside through one API. Each sandbox also has a 60fps desktop with Chromium, and real API keys stay outside it in a credentials proxy.
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
Dedalus Labs
Dedalus Labs is a Y Combinator company from San Francisco that sells Dedalus Machines, persistent Linux VMs for AI agents.
Each Dedalus Machine is a VM with its own Linux kernel, root and systemd, run on a Rust VMM forked from Cloud Hypervisor. You create a machine by size, run commands through an executions API or a temporary SSH session, and it sleeps after five idle minutes and wakes when the next command arrives. Only the root disk survives sleep, and machines get private IPs with no inbound connections.
- Machines per account
- 5 on Hobby, 20 on Pro, unlimited on Enterprise
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 | Dedalus Labs: price rank |
|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month A software factory needs to preview the apps it builds and branch from good states; boat has public HTTPS ports, snapshots and forks, and Dedalus machines take no inbound connections. | 3rd cheapest of 367 $634 a month | 252nd of 367, with limits $3,667 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, but on Dedalus it is reachable only through the API or SSH, and boat lets it host a webhook or dashboard. | 5th cheapest of 366 $657 a month | 271st of 366, with limits $4,715 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Dedalus suits short code bursts: machines sleep after five idle minutes, sleep is free, and the next command wakes them. | 186th of 345, with limits $2,000 a month | 122nd of 345, with limits $244 a month |
| Sandboxes for your users 300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month Giving each of your users a sandbox means many machines; Dedalus Pro stops at 20, while boat's paid plans start at 100 at once. | Cheapest of 361 $209 a month | 266th of 361, with limits $1,534 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 | 118th of 360, with limits $7,411 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 | 355th of 364, with limits $1,196 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month Both run Docker on a full VM; boat also has /dev/kvm for Android emulators and VMs, and Dedalus does not document nested virtualization. | 4th cheapest of 377 $86 a month | 174th of 377, with limits $724 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB Both keep a dev box's disk; boat adds HTTPS URLs for the apps you run on it, where Dedalus needs an SSH port forward that closes on sleep. | 18th cheapest of 368 $26 a month | 102nd cheapest of 368 $190 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 streamed at 60fps with Chromium for computer use; on Dedalus you would install and run your own display stack. | 6th cheapest of 369 $200 a month | 241st of 369, with limits $1,077 a month |
| Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month boat has Chromium and a desktop ready in every sandbox; Dedalus gives a plain Linux VM where you set up the browser yourself. | 7th cheapest of 367 $200 a month | 263rd of 367, with limits $537 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 | 265th of 346, with limits $363 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 | 192nd of 338, with limits $7,266 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 | 249th of 360, with limits $356 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 | Dedalus Labs |
|---|---|---|
| Isolation | Full VM with its own kernel on dedicated AMD Ryzen 9 9950X hosts | VM with its own kernel on a custom Cloud Hypervisor fork |
| Root, systemd and Docker | Yes, Docker preinstalled | Root and systemd; Docker not documented but a full Linux guest can run it |
| Nested virtualization | Yes, /dev/kvm for Android emulators and VMs | Not documented |
| Biggest machine | Up to 16 vCPU / 32 GB on the $100 plan and above | Up to 16 vCPU / 64 GiB on Pro, resizable after create |
| Inbound access | HTTPS URLs for up to 50 ports, token-protected by default; raw TCP; dedicated IPv4 or IPv6 | None; private IPs only, reach a machine through the API or an SSH port forward |
| Sleep and wake | Stop and start by API; a stopped sandbox costs nothing and keeps its disk | Sleeps after 5 idle minutes by default; an execution or SSH call wakes it |
| Snapshots | Disk saved every minute and on stop; restore any snapshot | No snapshot API; the root disk persists across sleep |
| Fork | Fork a running sandbox with its disk | No fork or clone operation |
| Custom image | Yes | No; machines are created by size only |
| Desktop and browser | Linux desktop streamed at 60fps, Chromium and computer use built in | Not provided; install your own GUI stack |
| Coding agents | Seven agents preinstalled, with a prompt, steer and events API | None inside; a skill file teaches your own agent to use the CLI |
| Keeping API keys out of the sandbox | Credentials proxy for model providers and any HTTPS API; switch a credential to another account when credits run out | Environment variables passed per execution |
| SSH | Yes | Temporary sessions that close after 30 idle minutes or 12 hours |
| Regions | EU only (Germany, Finland, France) | Not published; US company, no region option |
| 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 | Dedalus Labs |
|---|---|---|
| Full VM (own kernel) | Yes | Yes |
| 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 | Yes |
| Automatic snapshots | Yes | No |
| 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 | Yes |
| HTTPS preview URLs | Yes | No |
| Raw TCP inbound | Yes | Not published |
| 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 | No |
| GPU | No | No |
| Windows | No | No |
| Arm64 | No | No |
| Self-hosting | No | No |
| BYOC | No | Yes |
| Open source | No | No |
| Scoped API keys | Yes | Not published |
| Spend limits | Yes | Not published |
| Usage per sandbox (API) | Yes | Yes |
| Audit logs | No | Yes |
| SSO | No | Yes |
| SOC 2 | No | No |
| HIPAA | No | Not published |
| EU data residency | Yes | No |
| Zero data retention | Yes | Not published |
Performance
There is no comparable measurement for Dedalus Labs in the hpc-sandbox-benchmarks runs, so we can't compare CPU speed. boat did 26.69 runs per second on the Node.js build-and-test benchmark at 4 vCPU / 8 GB, on dedicated AMD Ryzen 9 9950X hosts.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Dedalus Labs 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 | Dedalus Labs, 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 | $50 · Hobby |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $26 · $20/mo | $199 · Hobby |
| 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 | $2,397 · Pro · 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 | $671 · Pro · 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 | $1,482 · Pro · with limits |
boat: Plans start at $20 a month, and the whole fee is usage credit. You pay per second while a sandbox runs, and a stopped sandbox costs nothing and keeps its disk.
Dedalus Labs: Dedalus bills about $0.045 per vCPU-hour and $0.015 per GiB of RAM per hour while a machine runs, and sleep is free. Each machine gets 10 GiB of disk free on Hobby and 20 GiB on Pro, extra disk costs $0.0001 per GiB-hour even while asleep, and Pro's $20 fee comes back as compute credit.
Limits in those scenarios
Dedalus Pro allows 20 machines, so the agent fleet, user sandboxes and burst scenarios all need its Enterprise plan, which has no public price. Machine creation also needs beta admission for your organization, and Dedalus does not publish an egress price. boat fits four of the five scenarios; for 1,000 short jobs at once, the $2,000 plan queues starts above 90 a minute.
| Scenario | boat | Dedalus Labs |
|---|---|---|
| One coding agent, workdays | None that bite |
|
| One agent that never stops | None that bite |
|
| 50 agents in parallel | None that bite |
|
| Sandboxes for your users | None that bite |
|
| 1,000 short jobs at once |
|
|
Which one should you use?
Use Dedalus Labs when
your agents run short bursts on private machines that should sleep between commands, need up to 64 GiB of RAM, or must run in your own cloud.
- Sleep when idle, wake on the next command. Good for agents that work in short bursts: idle machines sleep for free and wake when the next command or SSH call arrives.
- Big machines you can resize. Pro machines go up to 16 vCPU and 64 GiB of RAM, and you can change a machine's size after creating it.
- Enterprise in your own cloud. Enterprise teams can run machines in their own cloud, with SSO, RBAC and audit logs.
Use boat when
your agents build and run real things: apps they serve on public URLs, Docker and emulators, a browser on a desktop, with snapshots to roll back to and forks to try many ideas at once.
- Agents can host what they build. An agent that builds a web app, a webhook or a preview link on Dedalus has no URL to share. An SSH port forward is the workaround, and it closes on sleep and after 12 hours.
- Snapshots, forks and custom images. When an agent breaks its machine on Dedalus, there is no earlier state to go back to, and trying ten ideas from one good state means setting up ten machines from scratch.
- Agents, a desktop and keys built in. Dedalus gives a plain VM and passes secrets as environment variables, so you build the harness and the browser stack yourself, and the real keys sit where a prompt injection can read them.
Questions people ask
Is boat cheaper than Dedalus Labs?
In four of the five scenarios on this page, yes: a sandbox running all month is $26 on boat against $199 on Dedalus Hobby, and a fleet of agents is $317 against $2,397 on Pro. For 1,000 short jobs at once neither fits, since Dedalus Pro stops at 20 machines and boat's top plan queues starts.
Can I open a port or a preview URL on a Dedalus machine?
Not publicly. Dedalus machines get private IPs and don't accept inbound connections, so you reach them through the API or an SSH port forward, which closes when the machine sleeps. boat gives each sandbox HTTPS URLs for up to 50 ports, raw TCP and its own IP.
Does Dedalus keep my files when a machine sleeps?
Yes. The root filesystem survives sleep, but memory, running processes and /tmp do not, and there is no snapshot or fork API. boat also keeps the disk, saves a snapshot every minute, and can fork a running sandbox.
Can I run Docker on both?
Both are full VMs with root and systemd, so Docker can run on both. boat ships Docker preinstalled and also exposes /dev/kvm for Android emulators and VMs, which Dedalus does not document.
How many machines can I run on Dedalus Labs?
Hobby allows 5 machines and Pro allows 20; more needs Enterprise. boat's paid plans start at 100 sandboxes at once and go to 2,000, and you raise them yourself.
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