Railway vs Agent 37
Railway is an app platform. Its Sandboxes are Linux VMs that sit next to your apps and databases and are destroyed at an idle timeout unless you save a checkpoint; its Cloud Agents keep their disk. Agent 37 hosts one always-on agent per user, in a gVisor sandbox by default, that can sleep when idle, keep its memory and wake on the next message.
The companies and products
Railway
Railway is a cloud platform for deploying web apps, servers and databases. It also sells Sandboxes and Cloud Agents.
Railway's main product is app hosting: you deploy services and databases as containers on its own hardware. In 2026 it added Sandboxes: isolated Linux VMs that you create from the CLI or a TypeScript SDK, run commands in and destroy. It also added Cloud Agents: a personal VM for a coding agent that keeps its disk and sleeps or wakes. Cloud Agents are in Priority Boarding, and breaking changes may occur. Railway suits teams that already host on Railway and want sandboxes next to their apps.
Agent 37
Agent 37 hosts AI agents, one isolated, always-on sandbox per user, behind one API.
You create an instance from a template (Hermes, OpenClaw, Claude Code, Codex, OpenCode, Pi, Grok or your own Docker image) and talk to it through a chat API, files, exec, SSH and HTTPS URLs. The default instance is a container enforced by gVisor. A VM type gives a full Linux machine with its own kernel for Docker, Kubernetes and Android emulators. It is built for companies that give each of their customers an agent.
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. “28th cheapest of 365” means 27 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 | Railway: price rank | Agent 37: price rank | boat: price rank |
|---|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month In our default preset Agent 37 ranks 5th, boat 8th and Railway 13th. Railway keeps work through checkpoints you take or a Cloud Agent, and Docker there is yours to install and verify. | 28th cheapest of 365 $1,481 a month | 4th cheapest of 365 $551 a month | 5th cheapest of 365 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB Agent 37 ranks 1st in our default preset: each assistant can sleep when idle, keep its memory and wake on the next message. boat ranks 5th with auto-stop turned off. Railway ranks 23rd; an always-on assistant there needs the idle timeout set to 0 or a Cloud Agent, and memory bills the whole time. | 21st cheapest of 361 $3,977 a month | Cheapest of 361 $279 a month | 6th cheapest of 361 $657 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Agent 37 ranks 1st in our default preset with small default sandboxes. Railway ranks 34th, since it bills for CPU used and ends the sandbox when the job ends. boat's start limits make big bursts a compromise. | 42nd cheapest of 336 $413 a month | Cheapest of 336 $22 a month | 189th of 336, with limits $2,000 a month |
| Sandboxes for your users 300 at once · 2 vCPU / 4 GB · 30,000 sessions of 23 min a month Agent 37 ranks 1st and boat 2nd in our default preset. Agent 37 runs up to 1,000 awake instances after $1,000 of top-ups, not counting sleeping and stopped ones. Railway allows 100 sandboxes per environment on Pro. | 46th of 349, with limits $290 a month | Cheapest of 349 $79 a month | 3rd cheapest of 349 $209 a month |
| RL rollouts 1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month | 84th of 349, with limits $4,450 a month | 9th of 349, with limits $363 a month | 39th of 349, with limits $2,000 a month |
| Agent evals 250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month Agent 37 ranks 1st and boat 8th in our default preset. Railway can fork a running sandbox and boot many from one checkpoint, but caps how many run per environment. | 113th of 359, with limits $364 a month | Cheapest of 359 $58 a month | 10th cheapest of 359 $200 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month CI jobs that build containers need Docker. boat ranks 2nd in our default preset, with Docker on every sandbox. Agent 37 runs Docker on its VM type at 2.5x the compute rate, and Railway's sandbox docs do not cover Docker. | 116th cheapest of 374 $1,210 a month | 227th of 374, with limits $96 a month | 2nd cheapest of 374 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB | 69th cheapest of 371 $89 a month | 3rd cheapest of 371 $12 a month | 19th cheapest of 371 $26 a month |
| Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month In our default preset Agent 37 ranks 1st, boat 8th and Railway 13th. On Agent 37 you build a noVNC desktop template or use the Ubuntu VM template; on Railway you add Xvfb and noVNC with the template builder. boat streams a 60 fps desktop in every sandbox. | 47th cheapest of 367 $407 a month | 3rd cheapest of 367 $120 a month | 7th cheapest of 367 $200 a month |
| Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month | 53rd of 356, with limits $93 a month | 47th of 356, with limits $29 a month | 7th cheapest of 356 $200 a month |
| Scraping & bots (startup scale) 50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month | 65th cheapest of 343 $570 a month | Cheapest of 343 $32 a month | 6th cheapest of 343 $90 a month |
| Mass scraping & bots (undetectable) 1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month | 111th of 327, with limits $2,988 a month | 39th of 327, with limits $648 a month | 73rd of 327, with limits $1,800 a month |
| Parallel red teaming 200 at once · 2 vCPU / 4 GB · 5,000 sessions of 30 min a month Agent 37 ranks 1st and boat 5th in our default preset. Both fork a machine to try many paths at once; both forks copy the disk, not the memory. | 71st of 351, with limits $574 a month | Cheapest of 351 $43 a month | 3rd cheapest of 351 $200 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 | 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 | Not ranked: misses a requirement | Not ranked: misses a requirement |
Key features
| Feature | Railway | Agent 37 | boat |
|---|---|---|---|
| Kind of product | App platform for web apps and databases, with Sandboxes and Cloud Agents (Priority Boarding) | Hosting for AI agents, one always-on sandbox per user behind one API | Sandboxes for agents, each one a full VM |
| Isolation | Sandboxes and Cloud Agents are isolated Linux VMs; app services are containers | gVisor by default; a full VM as an opt-in instance type | Every sandbox is a full VM with its own kernel |
| When the sandbox goes idle | Sandboxes are destroyed after the idle timeout (30 minutes by default, or 0 for never); files survive in checkpoints you take. Cloud Agents sleep and keep their disk | Opt-in auto-sleep after a quiet period (default 15 minutes), memory checkpointed, wakes on the next request | Auto-stop on a timer from creation or resume (default 1 hour, not idle-based); the disk is kept at least 14 days, and a resume is like a reboot |
| Docker and /dev/kvm | Root shell on Ubuntu; Docker and /dev/kvm are not documented, so you install and verify them yourself | Docker and /dev/kvm on the VM type, at 2.5x the default compute rate, with no init system; not on the default gVisor type | Docker, root, systemd and /dev/kvm on every sandbox |
| Fork and snapshots | Fork a running sandbox's filesystem; named disk checkpoints boot new ones (disk only, no memory) | Fork any running, stopped or sleeping instance (disk copy, the fork boots fresh); nightly backups in 7 slots plus one on-demand, restored in place | Fork a running sandbox from its last 1-minute snapshot; disk saved every minute and at each stop; saved snapshots as templates |
| Custom images | Templates built from ordered steps (apt packages, env, commands) and cached; no custom Docker image for sandboxes | Your own linux/amd64 Docker image as a template | Start from boat's image or a saved snapshot; no custom Docker or OCI image |
| Coding agents | Claude Code, Codex, OpenCode and Pi in Sandboxes, run through exec; Cloud Agents keep terminal sessions you reattach to | Hermes, OpenClaw, Claude Code, Codex, Grok, OpenCode and Pi, with a chat API per instance | Seven coding-agent harnesses installed (Claude Code, Codex, OpenCode and others), with a prompt, steer and events API |
| Keeping keys out of the sandbox | Secrets go in as environment variables, or you sign in and save it in a checkpoint | Managed LLM, search and Composio through a per-instance token, metered at cost; your own keys go in as env vars | Optional AI credentials proxy ($1 a day plus $1 per 1,000 requests); with it on, real keys of supported providers stay out of the sandbox |
| Desktop and computer use | No desktop included; you add Xvfb and noVNC with the template builder. Chromium is in the Cloud Agent image | You set it up: a noVNC desktop through a template build, or the Ubuntu VM template | Built-in 1920x1080, 60 fps desktop and Chrome stream in every sandbox, with computer-use tools; embeddable stream |
| Reaching the sandbox | Up to 10 public HTTPS domains, set only at creation, with no built-in auth; SSH and port forwarding | HTTPS URL per port, signed URLs, public ports, custom domains; no public IP, SSH through an authenticated WebSocket tunnel | Up to 50 HTTPS URLs per sandbox, token-gated by default; public IPv6; the IP changes at each resume |
| CPU | Railway Metal, hardware Railway owns and runs | Shared by default; a Performance type on dedicated cores at 4x the compute rate | Shared vCPUs on AMD Ryzen hosts dedicated to boat |
| Biggest sandbox | 8 vCPU / 8 GB on Hobby; on Pro up to 32 vCPU / 32 GB per the docs (24 vCPU / 24 GB per the pricing page) | 4 vCPU on every type; disk up to 100 GB on the VM and Performance types | 8 vCPU / 16 GB on the $20 plan, 16 vCPU / 32 GB from $100 |
| Sandboxes at once | Per environment: 10 on Free, 50 on Hobby, 100 on Pro; Enterprise can ask for more. No published start limit | 3, 10, 50, 200 or 1,000 awake instances by lifetime top-ups; sleeping and stopped ones do not count | 100 to 2,000 at once by plan; included starts from 12 a minute and 200 a day ($20) to 90 and 2,400 ($2,000); extra starts $0.03 to $0.07 |
| Regions | US West, US East, Amsterdam and Singapore | Dedicated servers in the United States | Germany, Finland and France; no US or Asia region, and you cannot pick one. Snapshot copies can be stored outside the EU |
Every feature Battleships tracks (43)
| Feature | Railway | Agent 37 | boat |
|---|---|---|---|
| Full VM (own kernel) | Yes | No documented way | Yes |
| Docker inside | Not published | No documented way | Yes |
| Nested virtualization | Not published | No documented way | Yes |
| Root | Yes | Yes | Yes |
| Systemd | Not published | Not published | Yes |
| Desktop GUI | With your own setup | With your own setup | Yes |
| Browser | Yes | Yes | Yes |
| Computer-use API | No documented way | No documented way | Yes |
| SSH | Yes | Yes | Yes |
| Persistent disk | Yes | Yes | Yes |
| Pause/resume | Yes | Yes | Yes |
| Automatic snapshots | No documented way | Yes | Yes |
| Fork/clone | Yes | Yes | Yes |
| Live fork (no pause) | Yes | Yes | Yes |
| Memory fork | No documented way | No documented way | No documented way |
| Custom image (Docker or snapshot) | Yes | Yes | Yes |
| ≥24 h sessions | Yes | Yes | Yes |
| Idle auto-stop | Yes | Yes | No documented way |
| Wake on request | No documented way | Yes | No documented way |
| HTTPS preview URLs | Yes | Yes | Yes |
| Raw TCP inbound | No documented way | Not published | Yes |
| Public IPv4 | No documented way | No documented way | No documented way |
| Egress allowlist | No documented way | No documented way | No documented way |
| Secret proxy | No documented way | Yes | Yes |
| Secret proxy for any API | No documented way | No documented way | Yes |
| Preinstalled agents | Yes | Yes | Yes |
| Agent harness API | Yes | Yes | Yes |
| MCP server | Yes | No documented way | No documented way |
| GPU | No documented way | No documented way | No documented way |
| Windows | No documented way | No documented way | No documented way |
| Arm64 | No documented way | No documented way | No documented way |
| Self-hosting | No documented way | No documented way | No documented way |
| BYOC | Yes | No documented way | No documented way |
| Open source | No documented way | No documented way | No documented way |
| Scoped API keys | Not published | Not published | Yes |
| Spend limits | Yes | Yes | Yes |
| Usage per sandbox (API) | No documented way | Yes | Yes |
| Audit logs | Yes | Not published | No documented way |
| SSO | Yes | No documented way | No documented way |
| SOC 2 | Yes | Yes | No documented way |
| HIPAA | Yes | No documented way | No documented way |
| EU data residency | Yes | No documented way | No documented way |
| Zero data retention | Not published | Not published | Yes |
Performance
Railway and Agent 37 are not in the public benchmarks we use, so we cannot say which of the three 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; boat also places sandboxes on Ryzen 7 9700X hosts and does not promise a CPU model. Agent 37 says its default instances share host resources and sells a Performance type on dedicated cores.
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. Railway and Agent 37 are 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 | Railway, a month | Agent 37, a month | boat, 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 | $16 · Hobby · with limits | $6 · Wallet (any top-up) | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $70 · Hobby · with limits | $12 · Wallet (any top-up) | $26 · $20/mo |
| 50 agents in parallel 50 machines of 4 vCPU / 8 GB at once, 8-hour sessions, 1,100 sessions a month, Docker inside | $739 · Hobby · with limits | $329 · Wallet ($50 lifetime top-ups) | $317 · $20/mo × 3 seats |
| Sandboxes for your users 2 vCPU / 4 GB, 30,000 sessions of 10 minutes a month, up to 300 at once | $223 · Pro · with limits | $33 · Wallet ($1,000 lifetime top-ups) | $200 · $100/mo × 2 seats |
| 1,000 short jobs at once 2 vCPU / 4 GB, 1,000 at once, 3 minutes each, 200,000 jobs a month | $781 · Pro · with limits | $73 · Wallet ($1,000 lifetime top-ups) · with limits | $8,800 · $2000/mo · with limits |
Railway: Railway bills sandboxes and Cloud Agents per second at about $50 per vCPU-month on CPU actually used and $50 per GB-month on memory in use, plus $0.05 per GB out. Memory keeps billing while a sandbox waits, until it is destroyed. Plans: Free ($0 with $1 of usage a month), Hobby ($5) and Pro ($20); the Hobby and Pro fees count toward usage. There is no published price for sandbox disk, so our bills leave disk out.
Agent 37: Agent 37 uses a prepaid wallet with no plan fee, metered per minute: $0.80 per vCPU and $0.70 per GB of RAM per month while running, and $0.09 per GB of disk per month even while stopped or asleep. Performance compute costs 4x and VM compute 2.5x the default rate. New accounts get a one-time $5 credit after adding a card.
Agent 37: Plans cost $20, $100, $500 or $2,000 a month, and the whole plan price comes back as VM time for that month; unused included time expires. A default sandbox (4 vCPU / 8 GB) costs $0.036 an hour, billed per second. A stopped sandbox has no compute charge, but saved snapshots above the included count and the AI credentials proxy still bill. A card is required, also for the 7-day trial.
Limits in those scenarios
Railway is marked as not a clean fit in every scenario because it is an app platform first and does not price sandbox disk. Its cap of 100 sandboxes per environment on Pro means 300 sandboxes for your users or 1,000 jobs at once need several environments or an Enterprise limit. Agent 37 runs up to 1,000 awake instances after $1,000 of top-ups; every type stops at 4 vCPU, and the 1,000-job burst runs on shared vCPUs, which the table marks as a compromise. The 50-agent fleet needs Docker, so Agent 37 is priced on its VM type at 2.5x compute. On boat, the 1,000-job burst needs about 6,667 starts a day, above the 2,400 that the $2,000 plan includes, so extra starts are paid; starts above 270 a minute fail with 429, and boat never queues a start.
| Scenario | Railway | Agent 37 | boat |
|---|---|---|---|
| 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 |
|
|
|
Which one should you use?
Use Railway when
your app and databases already live on Railway and you want sandboxes on the same private network, many short jobs billed for CPU used, or regions in the US or Asia.
- Your apps and databases in one place. Useful if your app already runs on Railway: a sandbox can join the private network of your Postgres and services.
- Bigger machines and short jobs billed for CPU used. Useful for heavy or quick work: Railway sandboxes go up to 32 vCPU on Pro and bill for CPU actually used.
- Regions in the US, EU and Asia. Useful if your users are spread out: Railway runs in California, Virginia, Amsterdam and Singapore.
Use Agent 37 when
you host one mostly idle assistant per customer in the US that should sleep with its memory kept and wake on the next message, at the lowest price per hour.
- Sleeps when idle and keeps its memory. Useful for assistants that sit idle most of the day: an instance sleeps, keeps its memory and wakes on the next message.
- Bring your own Docker image. Useful if you already ship a Dockerfile: Agent 37 uses your own image as a template.
- Lower price per hour on its default type. Useful for many small agents: Agent 37's bill is the lowest of the three in four of our five scenarios.
Where boat fits
If your agents do real work for hours, like installs, builds, tests, Docker and browsers, look at boat. Every sandbox is a full VM with systemd that keeps its disk, can be forked while it runs, has a 60 fps desktop, and runs coding agents you drive from outside. boat runs only in Germany, Finland and France, cannot sleep on idle, and limits starts per plan.
Docker on every machine
Railway's sandbox docs do not cover Docker or /dev/kvm, so you install and verify them yourself. Agent 37's default gVisor type cannot run a Docker daemon, and its VM type costs 2.5 times as much compute and has no init system. Every boat sandbox has Docker, systemd and /dev/kvm.
A disk that stays
A Railway sandbox is destroyed after 30 idle minutes unless you set the timeout to 0 or take checkpoints, and memory bills until it is destroyed. Agent 37 keeps the disk and backs it up nightly. boat saves the disk every minute and at each stop, and keeps a stopped sandbox's disk for at least 14 days.
Agents driven from outside, keys kept outside
On both, your own API keys sit inside the machine as env vars. Agent 37's managed LLM token is an option for its own models. boat's optional paid credentials proxy keeps real keys of supported providers out, and its prompt, steer and events API replaces parsing exec output.
Use boat when your agents work for hours or days: a full VM with Docker, systemd and a desktop that keeps its disk, forks while running, and runs coding agents you drive from outside, within its per-plan start limits. Compare it one to one: boat vs Railway and boat vs Agent 37.
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
Questions people ask
Is Railway cheaper than Agent 37?
No. Agent 37 comes out lower in every scenario we price. One agent running all month is $70 on Railway against $12 on Agent 37, and 50 agents in parallel are $739 against $329, with boat at $317. Railway's numbers leave out disk, which it does not price.
Which is faster, Railway or Agent 37?
Neither is in the public benchmarks we use, so we cannot say. boat did 28.06 runs/s on the V8 Web Tooling Benchmark at 4 vCPU / 8 GB in StarSling's run of 3 October 2026, on a Ryzen 9 9950X host. Agent 37 sells a Performance type on dedicated cores if you need steady speed.
Can Railway or Agent 37 run Docker?
Railway's sandbox docs do not cover Docker, so you install and verify it yourself. Agent 37 runs Docker and Compose on its VM type, which bills compute at 2.5x, but not on its default gVisor type. On boat, Docker works in every sandbox.
What happens when a sandbox goes idle?
Railway destroys a sandbox after the idle timeout, 30 minutes by default, and your checkpoints survive; a Cloud Agent sleeps and keeps its disk. Agent 37 can sleep an instance after a quiet period, keep its memory and wake it on the next request. boat's auto-stop runs on a timer (default 1 hour), not on idle; a stopped sandbox keeps its disk for at least 14 days, and a resume is like a reboot.
Can they fork a machine?
All three can. Railway forks a running sandbox's filesystem. Agent 37 copies the disk of a running, stopped or sleeping instance. boat forks a running sandbox from its last 1-minute snapshot. None copies memory, so the fork boots fresh.
Where do Railway and Agent 37 run?
Railway has regions in California, Virginia, Amsterdam and Singapore. Agent 37 runs on dedicated servers in the United States. boat runs on machines in Germany, Finland and France; you cannot pick a region, and snapshot copies can be stored outside the EU.
Try boat
Install the CLI, then run boat new.
7-day free trial