Paperspace vs Cerebrium
Paperspace rents whole Linux VMs with Nvidia GPUs by the hour, and you work inside them over SSH. Cerebrium runs your model or app as containers behind HTTP or WebSocket endpoints that scale up with requests and back down to zero.
The companies and products
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
Cerebrium
Cerebrium is a serverless GPU cloud for real-time AI: voice agents, video models, LLMs and other ML apps.
You describe an app in a cerebrium.toml file, with your Python code or your own Dockerfile, and Cerebrium builds it and serves it as HTTP or WebSocket endpoints. It starts containers when requests arrive, scales them across GPUs and regions, and scales them back to zero when traffic stops. It is built for teams that serve models in production, not for agents that need a machine of their own, so this page compares the places where the two overlap.
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. “58th cheapest of 364” means 57 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 | Paperspace: price rank | Cerebrium: price rank | boat: price rank |
|---|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month A software factory needs Docker and a whole disk that lasts; Paperspace has both but is set up by hand over SSH, and Cerebrium has neither documented. | 141st of 367, with limits $1,261 a month | 182nd of 367, with limits $2,008 a month | 3rd cheapest of 367 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB | 179th of 366, with limits $1,560 a month | 221st of 366, with limits $2,534 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 Short code runs suit Cerebrium, which starts a container per request and bills per second, while Paperspace bills a full hour for each short run. | 179th of 345, with limits $1,451 a month | 225th of 345, with limits $131 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 is built to hand sandboxes to your users: Paperspace has no API to run code inside a machine, and Cerebrium gives stateless endpoints; boat gives each user a full VM. | 124th of 361, with limits $1,243 a month | 230th of 361, with limits $821 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 Cerebrium's per-second billing and fast scaling suit RL rollouts far better than Paperspace's hourly billing; boat ranks higher than both. | 193rd of 360, with limits $30,570 a month | 72nd of 360, with limits $4,068 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 | 58th cheapest of 364 $1,241 a month | 32nd cheapest of 364 $639 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 run CI jobs, but builds that need Docker inside work on a Paperspace VM and are not documented on Cerebrium. | 101st cheapest of 377 $1,026 a month | 81st cheapest of 377 $766 a month | 4th cheapest of 377 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB Paperspace fits a dev box that never stops, with SSH and a disk that keeps your installs; Cerebrium has no SSH and resets everything outside /persistent-storage. | 49th cheapest of 368 $63 a month | 259th of 368, with limits $102 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 has a Linux desktop for computer use: Paperspace streams desktops only on Windows machines new users cannot create, and Cerebrium has no desktop product; boat streams one at 60fps. | 284th of 369, with limits $1,655 a month | 179th of 369, with limits $575 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 | 319th of 367, with limits $1,241 a month | 221st of 367, with limits $287 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 | 192nd of 346, with limits $112 a month | 223rd of 346, with limits $195 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 | 92nd of 338, with limits $2,242 a month | 135th of 338, with limits $3,991 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 | 207th of 360, with limits $214 a month | 197th of 360, with limits $191 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 Both fit GPU eval jobs: Paperspace rents GPUs up to H100 by the hour, Cerebrium bills GPUs up to B200 per second, and boat has no GPUs. | 30th cheapest of 98 $2,982 a month | 28th cheapest of 98 $1,975 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 | Paperspace | Cerebrium | boat |
|---|---|---|---|
| Kind of product | GPU and CPU VM cloud for machine learning | Serverless GPU cloud that serves your app as autoscaling endpoints | Sandbox cloud for agents: VMs you create, fork and drive from code |
| Isolation | Full VM with its own kernel | Containers built from your code or Dockerfile; the isolation layer is not named | Full VM with its own kernel on dedicated AMD Ryzen 9 9950X hosts |
| Shell access | SSH into every machine | No SSH; you reach the app through its HTTP or WebSocket endpoints | SSH, root and systemd, plus an API, SDK and CLI to run commands and move files |
| Docker and nested virtualization | Docker runs inside; nested virtualization is not documented | The app is the container; Docker inside and /dev/kvm are not documented | Docker, systemd and /dev/kvm on every machine |
| Files after a stop | Whole disk kept, and billed while the machine is off | Only the shared /persistent-storage volume survives; files elsewhere reset | Whole disk saved every minute and on stop; a stopped sandbox costs nothing |
| Snapshots and copies | Auto-snapshots at most hourly; copies via a custom template from a stopped machine | No fork; an early-beta memory checkpoint speeds up new containers of the same build | Fork a running sandbox in one call |
| Scale to zero | No; a machine bills by the hour while it is on | Yes; apps scale to zero and start on the next request | No; a stopped sandbox is started by API |
| GPUs | Nvidia M4000 to H100, up to 8 per machine; H100 needs approval | T4 to B200 and RTX PRO 6000, up to 8 per container, plus Inferentia2 and Trainium | None, CPU only |
| Billing step | Per hour | Per second for vCPU, memory and GPU | Per second |
| Inbound access | A public IPv4 on every Linux machine, any port; no managed HTTPS URLs | One endpoint per app on one port; custom domains on Standard | HTTPS URLs for up to 50 ports, token-protected; raw TCP; dedicated IPv4 or IPv6 |
| Desktop and computer use | Desktop streaming only on Windows machines, which new users cannot create | No desktop or browser product | 60fps Linux desktop with Chromium, computer use and browser control |
| Coding agents | None built in; ML templates such as ML-in-a-Box | None to run inside; an Agent Skill and docs MCP server help your agent deploy | Claude Code, Codex, OpenCode and more preinstalled, with a prompt, steer and events API |
| Keeping keys out of the machine | No credentials proxy; Gradient secrets are injected as environment variables | Secrets are exposed to the app as environment variables | AI credentials proxy for model providers and any HTTPS API you register |
| Regions | New York, Amsterdam and Canada | US East, US Central, Stockholm and Finland; 8 more on request | EU only: Germany, Finland, France |
Every feature Battleships tracks (43)
| Feature | Paperspace | Cerebrium | boat |
|---|---|---|---|
| Full VM (own kernel) | Yes | No | Yes |
| Docker inside | Yes | No | Yes |
| Nested virtualization | Not published | No | Yes |
| Root | Not published | 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 | No | Yes |
| Persistent disk | Yes | Yes | Yes |
| Pause/resume | Yes | No | Yes |
| Automatic snapshots | No | No | Yes |
| Fork/clone | No | 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 | Yes | Yes | No |
| Wake on request | No | Yes | No |
| HTTPS preview URLs | No | Yes | Yes |
| Raw TCP inbound | Not published | Not published | Yes |
| Public IPv4 | Yes | No | No |
| Egress allowlist | No | 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 | No | No | No |
| GPU | Yes | Yes | No |
| Windows | No | No | No |
| Arm64 | No | 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 | Yes | Not published | Yes |
| Usage per sandbox (API) | No | Yes | Yes |
| Audit logs | Not published | Yes | No |
| SSO | Yes | No | 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 Paperspace or Cerebrium: neither was part of the public StarSling benchmark run, so we cannot say which is faster. boat did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB in that run. Both GPU clouds sell mainly on GPU speed, which this test does not cover.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Paperspace and Cerebrium 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 | Paperspace, a month | Cerebrium, 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 | $19 · Core usage | $53 · Hobby · with limits | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $63 · Core usage | $210 · Hobby · with limits | $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 | $804 · Core usage | $2,557 · Hobby · with limits | $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 | $1,241 · Core usage | $715 · Hobby | $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 | $8,274 · Core usage | $1,681 · Standard | $2,000 · $2000/mo · with limits |
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 and public IP keep billing until you delete the machine, even when it is off.
Cerebrium: Cerebrium bills per second for each vCPU, GB of memory and GPU while a container runs. Hobby is free plus compute, Standard is $100 a month plus compute, and the protected tier costs 2x on all compute.
Limits in those scenarios
Cerebrium does not fit the coding agent, always-on and 50-agent scenarios: only /persistent-storage survives a stop, so installed packages reset, and the fleet also needs Docker inside, which is not documented. Paperspace fits them all, but it bills by the hour, so a 10-minute user session or a 3-minute job is billed as a full hour; that is why sandboxes for your users cost $1,241 there against $715 on Cerebrium. Cerebrium does not publish an egress price, so the scenarios count none. boat fits every scenario except 1,000 short jobs at once, where its start limits queue the burst.
| Scenario | Paperspace | Cerebrium | boat |
|---|---|---|---|
| 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 Paperspace when
you need GPUs or large CPU machines that a person sets up and keeps running over SSH, with the whole disk kept.
- A whole machine with SSH. For people who want a normal Linux server: SSH in, run Docker, and keep the whole disk between sessions.
- Big CPU machines. For heavy single jobs: Paperspace's C10 machine has 32 vCPU and 244 GB of memory, with its own public IPv4.
- Cheap long-running machines. For a machine that stays on, the hourly rate is low: one agent that never stops costs $63 a month, against $210 on Cerebrium.
Use Cerebrium when
you need GPUs to serve a model, voice or video app as an endpoint that scales with traffic and down to zero.
- Endpoints that scale to zero. For apps that wait for traffic: containers start on the next request, grow under load, and cost nothing when idle.
- Per-second billing for short work. For many short jobs: 1,000 short jobs at once costs $1,681 a month on Cerebrium, against $8,274 on Paperspace.
- Newer GPUs and more regions. For serving large models: GPUs up to B200 and H200, in the US and EU, with Asia and more on request.
Where boat fits
Paperspace and Cerebrium are GPU clouds, and boat has no GPUs, so model training and serving belong on one of them. If what you are really running is agents that need a machine of their own, boat gives each one a full VM that code can create, fork and drive, with agents and a desktop already inside.
Driven from code, with the whole disk kept
Paperspace's API cannot run commands inside a machine, so you build your own SSH layer. Cerebrium has no shell and keeps only /persistent-storage, so an agent's installs reset. boat runs commands by API and saves the whole disk every minute.
Docker, /dev/kvm and fork while running
Cerebrium does not document Docker inside, and neither documents /dev/kvm. Copying a Paperspace machine means stopping it and making a template. A boat sandbox has Docker and /dev/kvm, and forks while it runs.
Agents, a desktop and keys kept outside
Neither ships coding agents or a Linux desktop, so you install and wire both yourself, and both put secrets in environment variables an agent can read. boat preinstalls Claude Code and Codex, streams a 60fps desktop, and keeps keys in a credentials proxy.
Use boat when your agents need a machine of their own: a full VM with Docker, /dev/kvm, a desktop and coding agents, driven and forked from code. Compare it one to one: boat vs Paperspace and boat vs Cerebrium.
Questions people ask
Is Paperspace cheaper than Cerebrium?
For machines that stay on, yes: one agent that never stops costs $63 a month on Paperspace and $210 on Cerebrium. For short work, Cerebrium wins because it bills per second: 1,000 short jobs at once costs $1,681 against $8,274 on Paperspace. boat comes to $26 for the always-on agent.
Which is faster, Paperspace or Cerebrium?
We have no comparable CPU measurement for either one, so we cannot say. boat did 26.69 runs per second on the StarSling CPU test at 4 vCPU / 8 GB. GPU speed depends on the card you pick on each.
Can Paperspace or Cerebrium run Docker?
Docker runs inside a Paperspace machine, since it is a full VM. Cerebrium builds your app from a Dockerfile, but running Docker inside the app is not documented. A boat sandbox runs Docker, systemd and /dev/kvm.
What happens to my files when the machine stops?
Paperspace keeps the whole disk and keeps billing for it while the machine is off. Cerebrium keeps only the /persistent-storage volume, and the rest resets. boat saves the whole disk every minute and a stopped sandbox costs nothing.
Which has better GPUs?
Cerebrium offers newer cards, up to B200 and H200, while Paperspace goes up to H100, which needs approval. Both allow up to 8 GPUs at once. boat has no GPUs.
Can I SSH into Cerebrium like Paperspace?
No. Cerebrium has no SSH; you reach your app through its HTTP or WebSocket endpoints. Paperspace and boat both give you SSH into a full Linux VM.
Try boat
Install the CLI, then run boat new.
7-day free trial