Vultr vs Scaleway
Vultr and Scaleway both rent plain KVM virtual machines that you set up and run yourself. Vultr runs in 33 locations around the world, while Scaleway is a European cloud in Paris, Amsterdam, Warsaw and Milan with Arm64 machines and outbound firewall rules.
The companies and products
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.
Scaleway
Scaleway is a European cloud and AI provider, part of the iliad Group, that rents virtual machines, GPUs, storage and managed services.
Scaleway Instances are KVM virtual machines you pick from fixed shapes, from small shared-vCPU Development Instances to dedicated-vCPU and Arm64 types, running Linux or Windows Server. You get root, SSH, a public IP and security groups, and you install and run your own software next to Scaleway products like managed databases, Kubernetes, object storage and GPU Instances. It is general cloud infrastructure, 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. “47th cheapest of 364” means 46 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 | Vultr: price rank | Scaleway: price rank | bochka: 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 that starts empty, so you install the agents yourself; bochka comes with agents, snapshots every minute and stops that cost nothing. | 114th of 367, with limits $1,060 a month | 49th of 367, with limits $322 a month | 3rd cheapest of 367 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB Either can run an assistant around the clock, but on both you build the agent setup, the HTTPS endpoint and the backups yourself; bochka has them built in. | 144th of 366, with limits $1,000 a month | 90th of 366, with limits $208 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 All three are a compromise for short bursts: Vultr and Scaleway bill each started hour, and bochka's start limits queue big bursts. | 168th of 345, with limits $1,000 a month | 166th of 345, with limits $893 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 Vultr does not publish its instance limit and Scaleway allows 15 DEV1-M after identity checks, so 300 user sandboxes need a support ticket on both; bochka allows 100 to 2,000 by plan. | 98th of 361, with limits $899 a month | 84th of 361, with limits $738 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 | 179th of 360, with limits $20,000 a month | 176th of 360, with limits $18,254 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 | 47th cheapest of 364 $893 a month | 42nd cheapest of 364 $776 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 run Docker builds on a full VM and bill at least a full hour per machine, and on both you run the runner software and the clean-up yourself. | 91st cheapest of 377 $857 a month | 82nd cheapest of 377 $777 a month | 4th cheapest of 377 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB All three fit a 24/7 dev box: Vultr and Scaleway are classic always-on servers, and a bochka machine keeps its disk with snapshots every minute. | 28th cheapest of 368 $40 a month | 30th cheapest of 368 $41 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 desktop built in, so you install one with a browser and VNC yourself; bochka streams a 60fps desktop with Chromium and computer use. | 252nd of 369, with limits $1,190 a month | 94th of 369, with limits $168 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 | 300th of 367, with limits $894 a month | 173rd of 367, with limits $126 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 | 175th of 346, with limits $89 a month | 149th of 346, with limits $13 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 | 72nd of 338, with limits $1,818 a month | 34th of 338, with limits $253 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 Scaleway's security groups can limit outbound traffic by port and IP range, while Vultr's firewall filters incoming traffic only; bochka gives each attacker its own full VM with open internet. | 168th of 360, with limits $150 a month | 103rd of 360, with limits $29 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 | Can't run it | 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 | Can't run it |
Key features
| Feature | Vultr | Scaleway | bochka |
|---|---|---|---|
| What you get | A general KVM server you set up and run yourself | A general cloud VM you set up and run yourself | A sandbox API for agents, on full VMs |
| Isolation | Every instance is a KVM VM with its own kernel | Every Instance is a QEMU/KVM VM with its own kernel | Every sandbox is a full VM with its own kernel |
| CPU | Shared vCPU plans, or dedicated AMD EPYC vCPU on Optimized and VX1 plans; no comparable measurement | Shared vCPU on Development and Play types, dedicated vCPU on others; no comparable measurement | Dedicated AMD Ryzen 9 9950X (Zen 5) hosts; 26.69 runs/s at 4 vCPU / 8 GB |
| Docker and nested virtualization | Root and Docker; virtualization documented for VX1 and bare metal only | Root and Docker on Linux images; nested virtualization is not documented | Docker, root, systemd and /dev/kvm on every machine, documented |
| Stopping | A stopped instance is billed in full until you destroy it | A console or API power-off stops compute billing, but disks and IPs keep billing; standby bills as running | A stopped sandbox costs nothing and keeps its disk |
| Snapshots | Manual disk snapshots that can take up to 30 minutes; paid scheduled backups | Images and volume snapshots you create yourself; none automatic | Automatic snapshots every minute and on stop |
| Fork | Snapshot the disk, then deploy new instances from it (no memory) | Create an image, then boot new Instances from it (no memory) | Fork a running sandbox in one call (disk, no memory) |
| How many at once | Account limit not published; raised by asking support | Published quotas per shape, for example 10 DEV1-L after identity checks; more through support | 100 to 2,000 sandboxes by plan, raised self-serve |
| Reaching the machine | Public IPv4 on any port, one IPv4 included; HTTPS needs your own domain, proxy and certificates | Public IPv4 and IPv6 on any port, IPv4 billed on top; HTTPS needs your own proxy, certificates and DNS | HTTPS URLs for up to 50 ports, token-protected; raw TCP; own IP |
| Desktop and coding agents | Headless Linux; you install a desktop, VNC and agents yourself | None built in; you install a desktop, browser and agents yourself | 60fps desktop with Chromium and computer use; Claude Code, Codex, OpenCode and more preinstalled, driven from outside by API |
| Keeping keys out of the machine | No credentials proxy; you put secrets on the server yourself | Secret Manager stores secrets that the machine fetches, so they end up on it | AI credentials proxy for model providers and any HTTPS API you register |
| Outbound rules | Vultr Firewall filters incoming traffic only; outbound rules go in the guest | Security groups with inbound and outbound rules by protocol, port and IP range | Open internet on every plan; no provider-side allowlist |
| GPUs, Windows and Arm | Nvidia and AMD GPUs; Windows Server for a license fee; no Arm64 | Nvidia GPU Instances billed per minute; Windows Server and Arm64 Instances | No GPUs; Linux x86 only, Wine for Windows apps |
| Regions | 33 locations on six continents | Paris, Amsterdam, Warsaw and Milan | EU only: Germany, Finland, France |
Every feature Battleships tracks (43)
| Feature | Vultr | Scaleway | bochka |
|---|---|---|---|
| Full VM (own kernel) | Not published | Yes | Yes |
| Docker inside | Yes | Yes | Yes |
| Nested virtualization | Not published | Not published | Yes |
| Root | Yes | Yes | Yes |
| Systemd | Yes | Yes | 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 | No | No | Yes |
| Automatic snapshots | No | No | Yes |
| Fork/clone | Yes | Yes | Yes |
| Live fork (no pause) | Not published | Yes | Yes |
| Memory fork | No | No | No |
| Custom image (Docker or snapshot) | Yes | Yes | Yes |
| ≥24 h sessions | Yes | Yes | Yes |
| Idle auto-stop | No | No | No |
| Wake on request | No | No | No |
| HTTPS preview URLs | No | No | Yes |
| Raw TCP inbound | Yes | Yes | Yes |
| Public IPv4 | Yes | Yes | No |
| Egress allowlist | No | Yes | 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 | Not published | No | No |
| GPU | Yes | Yes | No |
| Windows | Yes | Yes | No |
| Arm64 | No | Yes | 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 | Not published | No | Yes |
| Usage per sandbox (API) | No | No | Yes |
| Audit logs | Yes | Yes | No |
| SSO | Yes | Yes | No |
| SOC 2 | Not published | No | No |
| HIPAA | Not published | No | No |
| EU data residency | Yes | Yes | Yes |
| Zero data retention | Not published | Not published | Yes |
Performance
There is no comparable measurement for Vultr or Scaleway: neither was part of the public StarSling benchmark run where bochka did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB. Both sell shared and dedicated vCPU types, so results would depend on the 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 and Scaleway 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 | Vultr, a month | Scaleway, 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 | $10 · Usage account | $14 · Usage account | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $40 · Usage account | $41 · Usage account | $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 | $524 · Usage account | $537 · Usage account | $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 | $894 · Usage account | $734 · Usage account | $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 | $5,960 · Usage account | $4,892 · Usage account | $2,000 · $2000/mo · with limits |
Vultr: Each Vultr plan has a fixed hourly price, billed by the started hour, and most plans are capped at 672 hours a month. Billing runs from deploy to destroy, including while the instance is stopped, and one public IPv4 is included.
Scaleway: Each Scaleway Instance shape has an hourly price, billed per hour of uptime including startup and standby. Disks and public IPv4 bill on top, also while an Instance is powered off, and the table leaves them out.
Scaleway: 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
Vultr does not publish how many instances a new account can run; you ask support to raise the limit. Scaleway publishes quotas per shape, such as 10 DEV1-L and 15 DEV1-M after identity checks, with more only through support. So 50 agents, 300 user sandboxes and 1,000 short jobs at once fit on price on both but need a ticket first, and every short job pays a full hour. On bochka the limit is 100 to 2,000 sandboxes 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 | Vultr | Scaleway | 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 Vultr when
your users are outside Europe, you need AMD GPUs or Nvidia A100 and GH200 machines, or you want the lowest price for a server that runs only part of the day.
- 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.
- More kinds of GPUs. Useful for model work: Vultr rents Nvidia L40S, A100, H100, GH200 and B200 and AMD MI300X, MI325X and MI355X machines.
- Lower price for part-time machines. Useful for one agent on workdays: that comes to $10 a month on Vultr against $14 on Scaleway, with one IPv4 included.
Use Scaleway when
your machines must stay in a European cloud, you want outbound firewall rules or Arm64 Instances, or you need GPUs billed by the minute.
- A European cloud. Useful if your data must stay in the EU with a European provider: Scaleway runs in Paris, Amsterdam, Warsaw and Milan.
- Outbound firewall rules. Useful if agents must only reach some hosts: Scaleway security groups filter outbound traffic by protocol, port and IP range.
- Arm64 machines and clear quotas. Useful if you test on Arm64 or plan a fleet: Scaleway sells Arm64 Instances and publishes its per-shape quotas.
Where bochka fits
Vultr and Scaleway are general clouds, not agent sandboxes. If what you are really running is agents, each on its own Linux machine, bochka gives every agent a full VM that is ready to work, with coding agents, a desktop, HTTPS ports and snapshots every minute built in. It runs in the EU, like Scaleway, and starts hundreds of sandboxes without a quota ticket.
Free stops, and a disk saved every minute
Agents sit idle between tasks. A stopped Vultr instance bills in full, and a powered-off Scaleway Instance still bills its disk and IP. Snapshots on both are manual, so work since the last one is lost if an agent breaks the machine.
Agents, a desktop and keys handled for you
Both start as an empty OS image, so you install and supervise each agent, desktop and VNC yourself, and any API key you put on the machine is one the agent can read. bochka preinstalls agents, streams a 60fps desktop and proxies credentials.
Hundreds of machines with HTTPS, no ticket
Fifty agents or 300 user sandboxes are above the default limits on both, so you wait on support first. Then each machine gets a bare IP, and you set up your own domain, proxy and certificates to share a link.
Use bochka when each agent needs its own full VM that is ready to work: coding agents, a desktop, HTTPS ports and snapshots every minute built in, fork while running, stops that cost nothing, and hundreds of sandboxes in the EU without a quota ticket. Compare it one to one: bochka vs Vultr and bochka vs Scaleway.
Questions people ask
Is Vultr cheaper than Scaleway?
For part-time use, yes: one coding agent on workdays comes to $10 a month on Vultr against $14 on Scaleway, before Scaleway's disk and IPv4. Running 24/7 they are close, at $40 against $41, and 300 user sandboxes come to $734 on Scaleway against $894 on Vultr. On bochka those come to $20, $26 and $200.
Which is faster, Vultr or Scaleway?
There is no comparable measurement for either: neither was in the public StarSling benchmark run where bochka did 26.69 runs per second at 4 vCPU / 8 GB. Both sell shared and dedicated vCPU types, so speed depends on the plan.
Can Vultr and Scaleway run Docker?
Yes, both are full KVM VMs with root, so Docker works. Vultr documents virtualization support only for VX1 and bare metal, and Scaleway does not document nested virtualization. On bochka, Docker, systemd and /dev/kvm are documented on every machine.
Do I pay for a stopped machine?
On Vultr, yes, in full until you destroy the instance. On Scaleway, standby bills as running, and a console or API power-off stops compute but keeps billing disks and IPs. On bochka, a stopped sandbox costs nothing and keeps its disk.
Where do Vultr and Scaleway run?
Vultr has 33 locations on six continents, including several in Europe. Scaleway runs in Paris, Amsterdam, Warsaw and Milan. bochka runs only in the EU, in Germany, Finland and France, with EU data residency.
Is bochka SOC 2 compliant?
SOC 2 is in progress. We're open about where we stand: live status at status.bochka.xyz and our data retention policy at docs.bochka.xyz/data-retention.
Try bochka
Install the CLI, then run bochka new.
Product preview