Crusoe vs Paperspace
Both are GPU clouds that rent you plain Linux VMs you set up over SSH. Crusoe bills per second and goes up to Nvidia GB200 and AMD MI355X clusters with Kubernetes and Slurm, while Paperspace, now part of DigitalOcean, bills by the hour and offers a wide range of Nvidia cards from M4000 up to H100.
The companies and products
Crusoe
Crusoe calls itself "the AI factory company": it builds energy-first data centers and rents GPU and CPU machines on Crusoe Cloud.
Crusoe Cloud is infrastructure for training and serving models: Nvidia and AMD GPU VMs, general-purpose and storage CPU VMs, disks, managed Kubernetes and Slurm, and managed inference. You create Ubuntu VMs through a console, CLI, Terraform or API, log in over SSH and install what you need. It is not a sandbox product, so this page compares the CPU VMs where it overlaps with bochka.
Paperspace
Paperspace is a GPU cloud for machine learning, owned by DigitalOcean since 2023.
Paperspace Machines are full Linux VMs with a persistent disk, a public IP and optional Nvidia GPUs from M4000 up to H100, billed by the hour. Next to Machines it sells Gradient Notebooks, Deployments and Workflows for training and serving models. It is built for people who train, test and run models, not as a sandbox API: its API manages machines, and you work inside them over SSH.
- Operating systems
- Linux; Windows templates closed to new users since 1 July 2024
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. “43rd cheapest of 364” means 42 of the other 363 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 | Crusoe: price rank | Paperspace: price rank | bochka: price rank |
|---|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month Both give you a real VM with Docker, but no coding agents, no desktop and no fork of a running machine, so a fleet of coding agents is a compromise on both. | 232nd of 367, with limits $2,996 a month | 141st of 367, with limits $1,261 a month | 3rd cheapest of 367 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB | 234th of 366, with limits $3,000 a month | 179th of 366, with limits $1,560 a month | 5th cheapest of 366 $657 a month |
| Code interpreter bursts 100 at once · 1 vCPU / 2 GB · 200,000 sessions of 1 min a month Crusoe handles short bursts better than Paperspace because it bills per second and Paperspace bills each short run as a full hour. | 128th of 345, with limits $268 a month | 179th of 345, with limits $1,451 a month | 186th of 345, 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 Neither has HTTPS URLs or an API to run code inside a machine, so sandboxes for your users need your own layer on both; bochka has that built in. | 104th of 361, with limits $951 a month | 124th of 361, with limits $1,243 a month | Cheapest of 361 $209 a month |
| RL rollouts 1,000 at once · 2 vCPU / 4 GB · 1,000,000 sessions of 3 min a month | 71st of 360, with limits $4,055 a month | 193rd of 360, with limits $30,570 a month | 34th of 360, with limits $2,000 a month |
| Agent evals 250 at once · 4 vCPU / 8 GB · 15,000 sessions of 20 min a month Both fit agent evals as plain VMs, with Crusoe at $811 a month and Paperspace at $1,241, against $200 on bochka. | 43rd cheapest of 364 $811 a month | 58th cheapest of 364 $1,241 a month | 10th cheapest of 364 $200 a month |
| CI runners 40 at once · 8 vCPU / 16 GB · 6,000 sessions of 12 min a month Both can host a CI runner, and Crusoe's per-second billing suits short jobs better than Paperspace's hourly billing; bochka starts a runner per job and bills per second. | 48th cheapest of 377 $395 a month | 101st cheapest of 377 $1,026 a month | 4th cheapest of 377 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB Both run a dev box that never stops; Paperspace comes to $63 a month against $121 on Crusoe, and bochka to $26. | 87th cheapest of 368 $121 a month | 49th cheapest of 368 $63 a month | 18th cheapest of 368 $26 a month |
| Computer-use agents 50 at once · 4 vCPU / 8 GB · 20,000 sessions of 15 min a month Neither ships a Linux desktop: on Crusoe you install and stream one yourself, and Paperspace streams desktops only on Windows machines that new users cannot create; bochka has a 60fps desktop with Chromium. | 221st of 369, with limits $811 a month | 284th of 369, with limits $1,655 a month | 6th cheapest of 369 $200 a month |
| Browser-use agents 50 at once · 2 vCPU / 4 GB · 30,000 sessions of 10 min a month | 246th of 367, with limits $405 a month | 319th of 367, with limits $1,241 a month | 7th cheapest of 367 $200 a month |
| Scraping & bots (startup scale) 50 at once · 1 vCPU / 2 GB · 100,000 sessions of 3 min a month | 268th of 346, with limits $403 a month | 192nd of 346, with limits $112 a month | 6th cheapest of 346 $90 a month |
| Mass scraping & bots (undetectable) 1,000 at once · 1 vCPU / 2 GB · 3,000,000 sessions of 2 min a month | 201st of 338, with limits $8,055 a month | 92nd of 338, with limits $2,242 a month | 71st of 338, with limits $1,800 a month |
| Parallel red teaming 200 at once · 2 vCPU / 4 GB · 5,000 sessions of 30 min a month | 201st of 360, with limits $207 a month | 207th of 360, with limits $214 a month | 2nd cheapest of 360 $200 a month |
| GPU eval jobs 10 at once · 8 vCPU / 32 GB · 500 sessions of 30 min a month Paperspace is the better fit for GPU eval jobs: it rents single Nvidia cards by the hour, while Crusoe's H100, H200 and A100 SXM machines come only as 8-GPU VMs; bochka has no GPUs. | 86th of 98, with limits $7,803 a month | 30th cheapest of 98 $2,982 a month | 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 | Can't run it |
Key features
| Feature | Crusoe | Paperspace | bochka |
|---|---|---|---|
| Kind of product | GPU and CPU cloud for AI training, inference and HPC | GPU and CPU VM cloud for machine learning, owned by DigitalOcean | Sandbox cloud for agents: full VMs you create, fork and drive from code |
| Isolation | Full VM on cloud-hypervisor | Full VM with its own kernel | Full VM with its own kernel on dedicated AMD Ryzen 9 9950X hosts |
| GPUs | Nvidia L40S, A100, H100, H200, B200, B300, GB200 and AMD MI300X, MI355X; on-demand, spot or reserved | Nvidia M4000 up to A100 and H100, one to eight per machine; H100 needs approval | None |
| Billing step | Per second while running | Per hour while on | Per second while running |
| Stopped machines | No compute charge; the 128 GB OS disk keeps billing until you delete the VM | No compute charge; the disk and public IP keep billing until you delete the machine | A stopped sandbox costs nothing and keeps its disk |
| Docker and nested virtualization | sudo user; Docker preinstalled on GPU images; nested virtualization not documented | Docker runs inside; nested virtualization not documented | Root, systemd, Docker and /dev/kvm on every machine |
| Saving work | OS and attached disks survive a stop; no disk snapshots; local NVMe is erased on stop | Disk kept while off; auto-snapshots at most hourly, up to 10 kept | Automatic snapshots every minute and on stop |
| Copying a machine | Custom image of the boot disk, only while the VM is shut off | Custom template, only while the machine is off, same region only | Fork a running sandbox in one call; named snapshots as templates |
| Desktop and computer use | None; install and stream a desktop yourself | Desktop streaming only on Windows machines, closed to new users | 60fps desktop and Chromium, computer use and browser control |
| Coding agents | None; an MCP server inspects your Crusoe infrastructure | None built in; ML templates such as ML-in-a-Box | Claude Code, Codex, OpenCode and more preinstalled, driven from outside with a prompt, steer and events API |
| Reaching the machine | Public IPv4, SSH, TCP and UDP load balancers; no managed HTTPS URLs | Public IPv4 on every Linux machine, any port; no managed HTTPS URLs; no IPv6 | Token-protected HTTPS URLs for up to 50 ports, raw TCP, dedicated IPv6 or IPv4 |
| Keeping keys out of the machine | No credentials proxy; secrets live on the VM | No credentials proxy; Gradient secrets are environment variables | AI credentials proxy for model providers and any HTTPS API |
| Network rules | VPC firewall with allow rules by IP, port and protocol, in and out | No provider-side egress allowlist | Open internet on every plan; no provider-side allowlist |
| Regions | US (Texas, Virginia, Ohio, Nevada) and EU (Iceland, Norway) | New York, Amsterdam and Canada | EU only: Germany, Finland, France |
Every feature Battleships tracks (43)
| Feature | Crusoe | Paperspace | bochka |
|---|---|---|---|
| Full VM (own kernel) | Yes | Yes | Yes |
| Docker inside | Yes | Yes | Yes |
| Nested virtualization | No | Not published | Yes |
| Root | Yes | Not published | Yes |
| Systemd | Not published | Not published | Yes |
| Desktop GUI | No | No | Yes |
| Browser | No | No | Yes |
| Computer-use API | No | No | Yes |
| SSH | Yes | Yes | Yes |
| Persistent disk | Yes | Yes | Yes |
| Pause/resume | Yes | Yes | Yes |
| Automatic snapshots | No | No | Yes |
| Fork/clone | Yes | No | Yes |
| Live fork (no pause) | No | No | Yes |
| Memory fork | No | No | No |
| Custom image (Docker or snapshot) | Yes | Yes | Yes |
| ≥24 h sessions | Yes | Yes | Yes |
| Idle auto-stop | No | Yes | No |
| Wake on request | No | No | No |
| HTTPS preview URLs | No | No | Yes |
| Raw TCP inbound | Yes | Not published | Yes |
| Public IPv4 | Yes | Yes | No |
| Egress allowlist | Yes | No | No |
| Secret proxy | No | No | Yes |
| Secret proxy for any API | No | No | Yes |
| Preinstalled agents | No | No | Yes |
| Agent harness API | No | No | Yes |
| MCP server | Yes | No | No |
| GPU | Yes | Yes | No |
| Windows | No | No | No |
| Arm64 | Yes | No | No |
| Self-hosting | No | No | No |
| BYOC | No | No | No |
| Open source | No | No | No |
| Scoped API keys | Not published | Not published | Yes |
| Spend limits | No | Yes | Yes |
| Usage per sandbox (API) | Yes | No | Yes |
| Audit logs | Yes | Not published | No |
| SSO | Yes | Yes | No |
| SOC 2 | Yes | Yes | No |
| HIPAA | Yes | Yes | No |
| EU data residency | Yes | Yes | Yes |
| Zero data retention | Not published | Not published | Yes |
Performance
There is no comparable CPU measurement for Crusoe or Paperspace: neither was in the StarSling hpc-sandbox-benchmarks run we use. bochka did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB in that run, on dedicated AMD Ryzen 9 9950X (Zen 5) hosts. Crusoe's general-purpose CPU VMs run on AMD Genoa or Turin processors.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Crusoe and Paperspace have 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 | Crusoe, a month | Paperspace, a month | bochka, 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 | $32 · On-demand | $19 · Core usage | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $121 · On-demand | $63 · Core usage | $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 | $1,488 · On-demand | $804 · Core usage | $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 | $405 · On-demand | $1,241 · Core usage | $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 | $811 · On-demand | $8,274 · Core usage | $2,000 · $2000/mo · with limits |
Crusoe: Crusoe bills each VM per second while it runs, at a flat rate per vCPU that includes 4 GB of memory per vCPU, and every VM also pays for a 128 GB OS disk each month, running or stopped. It has no plan fee and does not charge for data transfer today; reservations bill for the whole term.
Paperspace: Paperspace bills each machine per hour while it is on, at a fixed rate per size (C5, 4 vCPU / 8 GB, is $0.08 an hour). The boot disk (from $5 a month for 50 GB) and the public IP keep billing until you delete the machine, even when it is off.
Paperspace: 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.
Limits in those scenarios
Crusoe sells CPU VMs with 4 GB of memory per vCPU, so the 4 vCPU rows use its 16 GB machine, which is why one agent that never stops comes to $121 a month on Crusoe and $63 on Paperspace. Paperspace bills a full hour for a short run, so 1,000 short jobs at once come to $8,274 a month there against $811 on Crusoe. Neither publishes start limits; both set per-account quotas that you raise through support. On bochka the 1,000-job burst is queued and needs more starts per day than the $2,000 plan allows.
| Scenario | Crusoe | Paperspace | bochka |
|---|---|---|---|
| One coding agent, workdays | None that bite | None that bite | None that bite |
| One agent that never stops | None that bite | None that bite | None that bite |
| 50 agents in parallel | None that bite | None that bite | None that bite |
| Sandboxes for your users | None that bite | None that bite | None that bite |
| 1,000 short jobs at once | None that bite | None that bite |
|
Which one should you use?
Use Crusoe when
you need the newest Nvidia or AMD GPUs in clusters, per-second billing for short jobs, Kubernetes or Slurm, or firewall rules for traffic in and out.
- The newest and biggest GPUs. For large training or inference runs, Crusoe rents Nvidia B200, B300 and GB200 and AMD MI300X and MI355X with InfiniBand.
- Per-second billing. For short or uneven jobs, Crusoe bills each VM per second while it runs, where Paperspace bills by the hour.
- Cluster tools and network control. For teams running clusters, Crusoe offers managed Kubernetes, Slurm, shared disks and VPC firewall rules in and out.
Use Paperspace when
you want one Nvidia GPU at a time by the hour, cheap CPU machines that stay on, or machines in New York, Amsterdam or Canada.
- Single GPUs of many sizes. For one-off model tests, Paperspace rents one Nvidia card at a time, from older M4000s up to A100 and H100.
- Lower rates for steady CPU machines. For a machine that stays on, Paperspace's C5 with 4 vCPU and 8 GB comes in well under Crusoe's 4 vCPU VM.
- Hourly snapshots and auto-shutdown. For people who forget machines, Paperspace can snapshot the disk every hour and shut a machine down after a set time.
Where bochka fits
Crusoe and Paperspace are GPU clouds, and bochka has no GPUs, so model training and GPU evals belong on one of them. If what you are really running is agents that write code, browse or use a computer, both give you a bare Linux VM you set up over SSH. bochka gives each agent a full VM that saves itself every minute, forks while running and comes with a desktop, coding agents and hosting ready.
Work that saves itself and forks live
Crusoe has no disk snapshots and Paperspace snapshots at most hourly, so an agent that breaks its machine can lose its work. To branch, both make you shut the machine off and make an image first.
Agents and a desktop inside
Neither ships coding agents or a Linux desktop, and neither API can run commands inside a machine. You would build the SSH layer, desktop stream and agent control yourself and keep them working.
Hosting and keys handled for you
Both give you a public IP but no HTTPS URLs, and keys sit on the VM where agent-written code can read them. bochka gives HTTPS on 50 ports, raw TCP and a credentials proxy for any HTTPS API.
Use bochka when your workload is agents rather than models: each one gets a full VM that saves itself every minute, forks while running, and comes with a desktop, coding agents, HTTPS hosting and a keys proxy. Compare it one to one: bochka vs Crusoe and bochka vs Paperspace.
Questions people ask
Is Crusoe cheaper than Paperspace?
For steady CPU machines, no: one agent that never stops comes to $121 a month on Crusoe and $63 on Paperspace, and 50 agents in parallel to $1,488 versus $804. For short runs Crusoe wins because it bills per second: sandboxes for your users come to $405 versus $1,241, and 1,000 short jobs at once to $811 versus $8,274.
Which has better GPUs, Crusoe or Paperspace?
Crusoe has the newer and larger machines: Nvidia B200, B300 and GB200 and AMD MI300X and MI355X, with H100, H200 and A100 SXM only as 8-GPU VMs. Paperspace rents single Nvidia cards from M4000 up to A100 and H100, which suits smaller tests. bochka has no GPUs.
Which is faster?
We have no comparable CPU measurement for either: neither Crusoe nor Paperspace was in the StarSling benchmark run we use. bochka did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB in that run.
Can Crusoe and Paperspace run Docker?
Yes. Both rent full VMs, and Docker runs inside; Crusoe preinstalls it on its GPU images. Neither documents nested virtualization, while bochka gives every sandbox root, systemd, Docker and /dev/kvm.
Are Crusoe or Paperspace sandboxes for AI agents?
No. Both are GPU clouds that rent plain Linux VMs you reach over SSH, with no desktop, built-in coding agents, credentials proxy or fork of a running machine. bochka is built for agents and has all of these.
Where do Crusoe and Paperspace run?
Crusoe runs in the US (Texas, Virginia, Ohio, Nevada) and in Iceland and Norway. Paperspace runs in New York, Amsterdam and Canada, with the full GPU range in New York. bochka runs only in the EU (Germany, Finland and France).
Try bochka
Install the CLI, then run bochka new.
Product preview