Google Compute Engine vs OVHcloud

Both rent general cloud VMs that you set up yourself. Google Compute Engine has more regions, more chip types and bigger GPUs, while OVHcloud is a French cloud whose shared Discovery instances cost less in every scenario we priced, but a turned-off instance keeps billing at the full rate.

The companies and products

Google Compute Engine

Google Compute Engine is Google Cloud's general-purpose VM service, not a sandbox product: you rent virtual machines by the second and set them up yourself.

You pick a machine type (E2, N2D, C3D, Arm T2A, GPU families and more), a boot image and a network, and Compute Engine boots a KVM-based VM with its own kernel and root access. Disks are Persistent Disk or Hyperdisk volumes, traffic goes through VPC firewall rules, and anything agent-specific (agents, desktops, HTTPS URLs, key handling) is yours to build. It is built for every kind of server workload, from websites to GPU training, across Google Cloud's worldwide regions.

Kind of product: General cloud VMs (a hyperscaler), not an agent sandbox
Isolation: Each instance is a full VM on Google's KVM-based hypervisor
Regions: 43 regions and 130 zones
Billing: Per second after a 1 minute minimum; a stopped VM stops CPU and memory charges, while disks and external IPs bill as long as they exist
Nested virtualization: Linux KVM only, and not on E2, memory-optimized, Arm or AMD VMs (except N4D); must be enabled per VM
Quota: Per-region CPU quotas that vary by project, plus a global CPU cap on some new projects; free trial accounts cannot request increases

OVHcloud

OVHcloud is a French cloud provider; its Public Cloud rents OpenStack virtual machines, storage and networking by the second.

A Public Cloud instance is a Linux or Windows VM picked from families such as General Purpose (B3), Compute (C3), Memory (R3), GPU and the shared-resource Discovery range (D2). You run your own software on it, next to OVHcloud products like private networks, gateways, load balancers, Kubernetes and object storage, through a Control Panel, an API or OpenStack tools. It is general infrastructure, not an agent sandbox, so this page compares the two where they overlap: giving an agent its own Linux machine.

Product: OpenStack instances: fully isolated VMs on shared physical hosts, with Linux and Windows Server images
Billing: Every second is billed after a 1-minute minimum; instances bill until you delete them
Pricing change: Since 2026-10-01, local storage and public IPv4 are billed on top of B3, C3 and R3 instances
Regions: Europe, North America (Canada and the US) and Asia-Pacific (Singapore, Sydney), plus Local Zones
Quotas: Raised in steps of 20, 50, 100 and 200 VMs by buying prepaid credit; more through support
GPUs: Nvidia L4 and L40S instances by the hour, among others
Certifications: Lists ISO 27001/27017/27018, SOC 1, 2 and 3, HDS, PCI DSS and HIPAA in its US datacentres

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. “16th cheapest of 364” means 15 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	Google Compute Engine: price rank	OVHcloud: price rank	bochka: price rank
Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month Both can host a software factory on plain VMs, but on either you build the agent setup, preview URLs and cloning yourself; bochka ships them on every machine.	120th of 367, with limits $1,106 a month	90th of 367, with limits $814 a month	3rd cheapest of 367 $634 a month
Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB	177th of 366, with limits $1,433 a month	136th of 366, with limits $934 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 bursts favor OVHcloud at $43 a month against $133 on Compute Engine, since both bill per second after a 1-minute minimum; bochka's start limits make big bursts a compromise.	92nd of 345, with limits $133 a month	56th of 345, with limits $43 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 OVHcloud comes out cheaper here but 300 sandboxes of 2 vCPU go past its largest self-serve quota, and Compute Engine may need a quota request too; bochka ranks first.	92nd of 361, with limits $836 a month	43rd of 361, with limits $299 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	36th of 360, with limits $2,085 a month	27th of 360, with limits $1,280 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 Compute Engine is a full match for agent evals and OVHcloud only a compromise; bochka also fits fully at $200 a month.	16th cheapest of 364 $263 a month	99th of 364, with limits $231 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-based CI runners on full VMs; jobs that need /dev/kvm are unreliable on OVHcloud and blocked on Compute Engine's E2 VMs, while bochka has /dev/kvm on every machine.	41st cheapest of 377 $359 a month	29th cheapest of 377 $300 a month	4th cheapest of 377 $86 a month
24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB A 24/7 dev box is a classic server on both, and OVHcloud's shared Discovery instance comes in at $34 a month against $109 on Compute Engine.	82nd cheapest of 368 $109 a month	23rd cheapest of 368 $34 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 or computer-use API, so you install and stream one yourself; bochka streams a 60fps desktop with Chromium out of the box.	129th of 369, with limits $293 a month	108th of 369, with limits $231 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	212th of 367, with limits $244 a month	174th of 367, with limits $128 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	218th of 346, with limits $179 a month	159th of 346, with limits $65 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	108th of 338, with limits $2,978 a month	47th of 338, with limits $1,290 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	152nd of 360, with limits $120 a month	120th of 360, with limits $64 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 rent Nvidia GPUs, with Compute Engine going up to H100, H200 and B200 and OVHcloud offering L4 and L40S; bochka has no GPUs.	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	Google Compute Engine	OVHcloud	bochka
What you get	A general cloud VM; you install and wire everything else	A general OpenStack VM you set up and run yourself	A sandbox API for agents, on full VMs with agents, desktop, URLs and snapshots set up
Isolation	Every instance is a full KVM-based VM with its own kernel	Every instance is a fully isolated VM with its own kernel	Every sandbox is a full VM with its own kernel
Docker and nested virtualization	Docker, root and systemd; nested KVM only on some Intel families and N4D, not on E2, Arm or other AMD VMs	Docker and root; the vmx flag is passed, but a live migration can make the guest kernel panic	Docker, root, systemd and /dev/kvm on every machine, documented
Stopping	A stopped VM stops CPU and memory charges; disks and external IPs keep billing	Turned-off and paused instances bill at the same rate; shelving frees compute and bills a disk snapshot	A stopped sandbox costs nothing and keeps its disk
Snapshots	On demand or on a schedule you set	On-demand instance backups, or a schedule such as daily backups keeping the last 7 or 14	Automatic snapshots every minute and on stop
Pause and fork	Suspend keeps memory for up to 60 days; cloning means a machine image and a new VM	No memory suspend; copying means a backup, then a new instance from it	Fork a running sandbox in one call (disk, no memory); resume reboots from disk
Coding agents	None preinstalled; you install agents and drive them over SSH	None preinstalled; you install agents and drive them over SSH	Claude Code, Codex, OpenCode and more preinstalled, with a prompt, steer and events API
Keeping keys out of the machine	No credential proxy; service account tokens and Secret Manager values are read from inside the VM	No credential proxy; secrets live on the instance or are fetched from inside it	AI credentials proxy for model providers and any HTTPS API; switch a credential to another account when credits run out
Desktop and computer use	No managed desktop, browser or computer-use API on Linux VMs	None built in; you install a desktop and connect over SSH or RDP	60fps desktop and Chromium, computer use and browser control, embeddable stream
Reaching the machine	External IPv4 on any port; HTTPS needs your own certificate and a load balancer or web server	Public IPv4 on any port; HTTPS needs your own certificates or a load balancer	HTTPS URLs for up to 50 ports, token-protected; raw TCP; own IP
Outbound network rules	VPC egress firewall rules by IP and port; domain rules on Cloud NGFW Standard, a paid tier	Security groups with outbound rules by IP range, port and protocol	Open internet on every plan; no provider-side allowlist
GPUs	Nvidia GPUs from T4 and L4 up to H100, H200, B200 and GB200	Nvidia GPU instances such as L4, L40S and V100S	None
Operating systems and CPUs	Linux and Windows Server on Intel, AMD and Arm	Linux and Windows Server images; no Arm instances	Linux on dedicated AMD Ryzen 9 9950X (Zen 5) hosts; Wine for Windows apps
Regions	43 regions and 130 zones worldwide	Europe, Canada, the US, Singapore and Sydney, plus Local Zones	EU only: Germany, Finland, France

Every feature Battleships tracks (43)

Feature	Google Compute Engine	OVHcloud	bochka
Full VM (own kernel)	Yes	Yes	Yes
Docker inside	Yes	Yes	Yes
Nested virtualization	No	Yes	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	Yes	Yes	Yes
Automatic snapshots	No	Yes	Yes
Fork/clone	Yes	Yes	Yes
Live fork (no pause)	Yes	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	Yes	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	No	No	No
GPU	Yes	Yes	No
Windows	No	Yes	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	Not published	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

Google Compute EngineNot measured

OVHcloudNot measured

bochka26.69 runs/s

There is no comparable measurement for Compute Engine or OVHcloud: neither was part of the public StarSling benchmark run we use. In that run, on 2026-09-29, bochka did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB. Speed on either cloud depends on the machine family you pick, and OVHcloud's Discovery (D2) instances share their CPU.

Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Google Compute Engine and OVHcloud 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	Google Compute Engine, a month	OVHcloud, 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	$31 · Pay-as-you-go billing account (no plan fee)	$8 · Usage account	$20 · $20/mo
One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month	$64 · Pay-as-you-go billing account (no plan fee) · with limits	$34 · 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	$1,291 · Pay-as-you-go billing account (no plan fee)	$406 · 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	$378 · Pay-as-you-go billing account (no plan fee)	$128 · Usage account · with limits	$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	$690 · Pay-as-you-go billing account (no plan fee)	$256 · Usage account · with limits	$2,000 · $2000/mo · with limits

Google Compute Engine: Compute Engine bills each VM per second after a 1 minute minimum, here E2 sizes on demand in us-central1, with no platform fee. Disks and external IPs bill on top, also while stopped, and the lowest always-on rate needs a 3-year commitment.

OVHcloud: OVHcloud bills every second an instance exists after a 1-minute minimum, until you delete it; turned-off and paused instances bill at the same rate. The prices here use shared Discovery d2 instances, and since October 2026 dedicated B3, C3 and R3 instances also bill local storage and a public IPv4.

OVHcloud: 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

Compute Engine limits vCPUs per region with quotas that vary by project, and free trial accounts cannot ask for more, so 50 agents or 300 user sandboxes may need a quota request first. One agent that never stops reaches the $64 a month shown only with a 3-year commitment. OVHcloud raises quotas in steps of 20, 50, 100 and 200 VMs by buying prepaid credit, so 300 user sandboxes or 1,000 jobs at once are not confirmed, and its shared Discovery CPUs can slow heavy work. On bochka, the 1,000-job burst is queued: the $2,000 plan allows 90 starts a minute, 600 an hour and 2,400 a day.

Scenario	Google Compute Engine	OVHcloud	bochka
One coding agent, workdays	None that bite	None that bite	None that bite
One agent that never stops	3-year commitment	None that bite	None that bite
50 agents in parallel	None that bite	how many machines can run at once: limit not published, probably fine at 50 (the typical published limit is 50)	None that bite
Sandboxes for your users	None that bite	not confirmed: how many machines can run at once (you need 300; the typical published limit is 50)	None that bite
1,000 short jobs at once	None that bite	not confirmed: how many machines can run at once (you need 1,000; the typical published limit is 50) shared vCPUs (no guaranteed share of a core): heavy, sustained CPU use can be slowed by other tenants	max 2,400 starts/day (needs ~6,667) max 600 starts/hour (needs ~1,000) bursts above 90 starts/min are queued (a burst of 333 takes ~4 min to start)

Which one should you use?

Use Google Compute Engine when

you need the largest GPUs, Arm machines, regions far outside Europe, VMs that suspend with memory, or domain-based egress rules, and you have people to build the tooling.

Biggest GPUs, Arm and Windows. Useful if your agents need top GPUs or Arm: Compute Engine rents Nvidia H100, H200 and B200 and has Arm machine types.
Regions all over the world. Useful if users sit in Asia, the Americas, Africa or Australia: Compute Engine runs in 43 regions and 130 zones.
Cheaper stops and memory suspend. Useful for VMs that idle: a stopped VM bills only disks and IPs, and suspend keeps memory for up to 60 days.

Use OVHcloud when

you want lower list prices from a French cloud, scheduled backups, Windows or L4 and L40S GPUs, and long-lived servers that you delete or shelve when idle.

Lower list prices. Useful on a tight budget: OVHcloud's shared Discovery instances came out cheaper than Compute Engine in every scenario we priced.
Scheduled backups built in. Useful if you want backups without writing tooling: OVHcloud can run daily instance backups and keep the last 7 or 14.
A French cloud with GPUs and Windows. Useful for European teams that want Windows Server or Nvidia L4 and L40S GPUs from a provider based in France.

Where bochka fits

Compute Engine and OVHcloud are general clouds, and bochka is not trying to be one: it has no GPUs, no Windows and runs only in the EU. If what you are really running is agents, bochka gives each one a full VM that is ready to work, so you skip the agent setup, desktop, URLs and backup tooling you would build on either cloud.

/dev/kvm on every machine

Agents that run Android emulators or build VM images need KVM. Compute Engine blocks it on E2, Arm and most AMD VMs, and OVHcloud warns a live migration can crash nested guests. bochka documents /dev/kvm everywhere.

Snapshots every minute, free stops and fork

On OVHcloud a turned-off instance bills at the full rate and backups run daily, and on Compute Engine disks bill while stopped and a copy needs a machine image. bochka saves disk every minute and forks a running sandbox.

Agents, desktop and HTTPS ready

Both clouds hand you an empty OS image, so you install each agent, a desktop and certificates, and keys sit where the agent can read them. bochka ships coding agents with an API, a 60fps desktop, HTTPS on 50 ports and a credentials proxy.

Use bochka when what you run is agents: each gets a ready full VM with coding agents and an API, keys kept outside, /dev/kvm, a desktop, HTTPS URLs, snapshots every minute, fork while running and stops that cost nothing. Compare it one to one: bochka vs Google Compute Engine and bochka vs OVHcloud.

Questions people ask

Is OVHcloud cheaper than Google Compute Engine?

In our scenarios, yes. 50 agents in parallel come to $406 a month on OVHcloud against $1,291 on Compute Engine, and 300 user sandboxes to $128 against $378, though OVHcloud's quota for that many is not confirmed. bochka comes to $317 and $200 for the same two.

Which is faster, Compute Engine or OVHcloud?

We have no comparable measurement for either. Both depend on the machine family you choose, and OVHcloud's cheapest Discovery instances share their CPU. bochka did 26.69 runs per second at 4 vCPU / 8 GB in the StarSling benchmark.

Can Compute Engine and OVHcloud run Docker and nested VMs?

Both run Docker, since each instance is a full VM with root. Compute Engine allows nested virtualization only on some Intel families and N4D, not on E2, Arm or other AMD VMs. OVHcloud passes the vmx flag, but warns a live migration can make the guest kernel panic.

Do I pay for a stopped VM?

On Compute Engine, a stopped VM stops CPU and memory charges but its disks and external IPs keep billing. On OVHcloud, turned-off and paused instances bill at the same rate until you delete or shelve them. On bochka, a stopped sandbox costs nothing and keeps its disk.

Which has more regions?

Compute Engine runs in 43 regions and 130 zones worldwide. OVHcloud runs in Europe, Canada, the US, Singapore and Sydney, plus Local Zones. 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