Azure Virtual Machines vs IONOS Cloud

Both rent plain cloud VMs with their own kernel that you set up yourself. Azure is Microsoft's hyperscaler, with 77 regions and hundreds of sizes. IONOS Cloud is a German host with servers in Europe and the US, per-minute Cubes and 2 TB of free outgoing traffic a month.

The companies and products

Azure Virtual Machines

Azure Virtual Machines is Microsoft's general-purpose cloud VM service: you rent Linux or Windows machines by the minute in Azure regions worldwide.

It is a hyperscaler IaaS product, not a sandbox API. Each VM runs on Microsoft's Hyper-V hypervisor with its own kernel, root and SSH. You pick from hundreds of sizes, including Arm, burstable and GPU families. Networking, disks, images, firewalls and agent tooling are separate pieces that you set up. It suits teams that already run on Azure and want full control.

IONOS Cloud

IONOS Cloud is the public cloud of IONOS, a German hosting company in the United Internet group. It is a hyperscaler-style IaaS cloud, not an agent sandbox.

You rent KVM virtual machines: Cubes (fixed sizes from 1 to 16 vCPU with a local NVMe disk, billed by the minute), vCPU servers up to 60 vCPU and 230 GB of RAM, Dedicated Core servers and H200 GPU VMs. You build them in the Data Center Designer or through the Cloud API, SDKs and Terraform, next to block storage, private networks, firewalls, backups and managed services. It is for teams that want to run their own servers in Germany, the rest of Europe or the US, and set up the software on them themselves.

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. “5th cheapest of 359” means 4 of the other 358 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 workloadAzure Virtual Machines: price rankIONOS Cloud: price rankboat: price rank
Software factory
100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month
114th of 365, with limits
$730 a month
99th of 365, with limits
$492 a month
5th cheapest of 365
$634 a month
Always-on AI employees
50 machines 24/7 · 2 vCPU / 4 GB
An always-on assistant fits on either cloud, but you set up the agent, its keys and its desktop yourself. boat ranks 5th.
176th of 361, with limits
$644 a month
174th of 361, with limits
$540 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
IONOS fits short code runs better than Azure (49th against 87th), if you can create and delete VMs quickly through its API.
118th of 336, with limits
$222 a month
61st of 336, with limits
$27 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
Neither cloud gives you a sandbox layer for your users, and IONOS does not publish how many machines can run at once. boat ranks 2nd here.
69th of 349, with limits
$500 a month
42nd of 349, with limits
$182 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
Both clouds do well for RL rollouts (IONOS 12th, Azure 20th), since large fleets of plain VMs are what they rent. boat is 29th.
22nd of 349, with limits
$1,039 a month
12th of 349, with limits
$740 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
Azure Virtual Machines ranks 3rd for agent evals, a full fit thanks to per-minute billing and many sizes. boat is 8th and IONOS 65th.
5th cheapest of 359
$173 a month
101st of 359, with limits
$137 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
IONOS Cloud ranks 1st for CI runners, with per-minute Cubes and no plan fee. boat is 2nd and Azure 26th.
33rd cheapest of 374
$335 a month
Cheapest of 374
$60 a month
2nd cheapest of 374
$86 a month
24/7 dev boxes
1 machine 24/7 · 4 vCPU / 8 GB
A dev box that stays on all month is a plain VM job. IONOS is 6th here, boat 13th and Azure 51st.
74th cheapest of 371
$94 a month
9th cheapest of 371
$20 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
On both clouds you set up the desktop yourself: install a desktop and an RDP, VNC or WebRTC server. boat ranks 8th, with a desktop on every sandbox.
143rd of 367, with limits
$192 a month
133rd of 367, with limits
$137 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
233rd of 356, with limits
$107 a month
226th of 356, with limits
$74 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
258th of 343, with limits
$116 a month
227th of 343, with limits
$40 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
84th of 327, with limits
$2,044 a month
42nd of 327, with limits
$897 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
164th of 351, with limits
$46 a month
160th of 351, with limits
$38 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 requirementNot ranked: misses a requirementNot 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 requirementNot ranked: misses a requirementNot ranked: misses a requirement

Key features

FeatureAzure Virtual MachinesIONOS Cloudboat
Kind of productGeneral cloud VMs; you set up the image, networking and toolingGeneral cloud servers (Cubes, vCPU and Dedicated Core servers) plus storage, networks and managed servicesSandboxes for AI agents: a full VM per sandbox, made and driven by API
IsolationFull Hyper-V VM with its own kernelKVM VM with its own kernelFull VM with its own kernel, on physical hosts dedicated to boat (AMD Ryzen 7 9700X and Ryzen 9 9950X); vCPUs are shared
Docker and nested virtualizationRoot and Docker on any size; nested virtualization on many Intel and AMD sizes (such as D, E and F series), not on Arm or B-seriesRoot Linux VM where you install Docker; IONOS documents nested Windows VBS and WSL 2 (slower, software emulation), and has no guide for KVM in a Linux guestDocker, root, systemd and /dev/kvm ready on every sandbox
Stopping a machineDeallocate stops compute billing; disks and static IPs keep billing; a shutdown from inside the OS keeps billing computeA suspended Cube releases CPU, RAM and IPs, but its storage keeps billing; a stopped vCPU server stops CPU and RAM billing; a $0 bill needs a snapshot and a deleteNo compute charge while stopped; disk kept at least 14 days, longer as a saved snapshot
SnapshotsDisk snapshots on demand; scheduled snapshots need Azure Backup, a separate paid serviceManual full snapshots of one volume, usable only in the same location; scheduled backups through the paid IONOS Backup ServiceAutomatic disk snapshots every minute and at each stop; writes after the last good snapshot can be lost if a machine fails
Fork and cloneCapture a gallery image (the VM stops briefly) or snapshot its disks, then create new VMs that boot coldSnapshot a disk of a running server, then create new VMs from it; disk only, no memoryFork a running sandbox in one call, from its last 1-minute snapshot (disk, no memory)
Pause with memoryHibernation saves memory to the OS disk on supported sizes (up to 64 GB of RAM on general-purpose sizes, 112 GB on some GPU sizes)Not documented; suspend turns the Cube offResume reboots from disk; memory and running programs are not saved
Desktop and computer useYou set it up: install a desktop such as xfce and xrdp, then open port 3389 (Microsoft guide)You set it up: install a desktop, a VNC or WebRTC server and a browser; the Remote Console is for troubleshooting1920x1080 desktop and Chrome at 60 fps, computer-use tools and browser control, embeddable stream
Coding agents and keysYou set it up: install agents with Run Command, scripts or SSH; Key Vault secrets are fetched into the VM, where code can read themYou set it up: install the agent CLIs and put keys into the VM, or run your own proxy; an MCP server lets your AI client manage IONOS resourcesSeven harnesses installed, including Claude Code, Codex and OpenCode, with a prompt, steer and events API; optional AI credentials proxy ($1 a day plus $1 per 1,000 requests)
Reaching the machinePublic IPv4, static if you pay for it, and any port through security group rules; you set up HTTPSPublic IPv4 on each VM and any port you open; reserved IPv4 at 5 EUR per 30 days; private LANs and load balancersUp to 50 HTTPS URLs per sandbox, token-gated by default; public IPv6; no static IPv4, and the IP changes at each resume
Outbound rulesSecurity groups by IP and port, plus domain allowlists with the paid Azure FirewallNetwork Security Groups with egress rules by IP, port and protocolOpen internet on every plan; no provider-side allowlist
GPUs, Windows and CPUsNvidia H100 and A100 GPU sizes; Windows and Linux on Intel, AMD and ArmNVIDIA H200 VMs with 1 to 8 GPUs; Windows Server and LinuxLinux x86_64 only; no GPU on any size; Windows apps through Wine
Regions77 regions worldwide, available or coming soonGermany, UK, France, Spain and the US; you pick the data centerGermany, Finland and France; you cannot pick a region; no US or Asia region
Every feature Battleships tracks (43)
FeatureAzure Virtual MachinesIONOS Cloudboat
Full VM (own kernel)YesYesYes
Docker insideYesYesYes
Nested virtualizationYesNot publishedYes
RootYesNot publishedYes
SystemdYesNot publishedYes
Desktop GUIWith your own setupNo documented wayYes
BrowserNo documented wayNo documented wayYes
Computer-use APINo documented wayNo documented wayYes
SSHYesYesYes
Persistent diskYesYesYes
Pause/resumeYesYesYes
Automatic snapshotsNo documented wayWith your own setupYes
Fork/cloneYesYesYes
Live fork (no pause)No documented wayYesYes
Memory forkNo documented wayNo documented wayNo documented way
Custom image (Docker or snapshot)YesYesYes
≥24 h sessionsYesYesYes
Idle auto-stopNo documented wayNo documented wayNo documented way
Wake on requestNo documented wayNo documented wayNo documented way
HTTPS preview URLsNo documented wayNot publishedYes
Raw TCP inboundYesNot publishedYes
Public IPv4YesYesNo documented way
Egress allowlistYesYesNo documented way
Secret proxyNo documented wayNo documented wayYes
Secret proxy for any APINo documented wayNo documented wayYes
Preinstalled agentsNo documented wayNo documented wayYes
Agent harness APINo documented wayNo documented wayYes
MCP serverYesYesNo documented way
GPUYesYesNo documented way
WindowsYesYesNo documented way
Arm64YesNo documented wayNo documented way
Self-hostingNo documented wayNo documented wayNo documented way
BYOCNo documented wayNo documented wayNo documented way
Open sourceNo documented wayNo documented wayNo documented way
Scoped API keysNot publishedNot publishedYes
Spend limitsNo documented wayNot publishedYes
Usage per sandbox (API)YesYesYes
Audit logsYesNot publishedNo documented way
SSOYesYesNo documented way
SOC 2YesNot publishedNo documented way
HIPAAYesNo documented wayNo documented way
EU data residencyYesYesNo documented way
Zero data retentionNot publishedNot publishedYes

Performance

Azure Virtual MachinesNot measured
IONOS CloudNot measured
boat28.06 runs/s

Azure Virtual Machines and IONOS Cloud are not in the public benchmarks we use, so we cannot say which is faster. Both offer many CPU types that vary by size or location. In StarSling's run of 3 October 2026 (run 37073680030), boat did 28.06 V8 Web Tooling Benchmark runs per second 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, which that run did not test, and does not promise a CPU model.

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. Azure Virtual Machines and IONOS Cloud 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.

ScenarioAzure Virtual Machines, a monthIONOS Cloud, a monthboat, 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
$26 · Pay-as-you-go subscription (no plan fee)$5 · Pay as you go$20 · $20/mo
One agent that never stops
4 vCPU / 8 GB, 50 GB disk, running 24/7 all month
$48 · Pay-as-you-go subscription (no plan fee) · with limits$20 · Pay as you go$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
$1,168 · Pay-as-you-go subscription (no plan fee)$240 · Pay as you go$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
$225 · Pay-as-you-go subscription (no plan fee)$74 · Pay as you go · 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
$483 · Pay-as-you-go subscription (no plan fee) · with limits$148 · Pay as you go · with limits$8,800 · $2000/mo · with limits

Azure Virtual Machines: Azure bills each VM per full minute while it runs, with no plan fee. Disks, public IPv4 addresses and egress past 100 GB a month bill on top, and disks keep billing while the VM is deallocated. The always-on price in the table needs a 3-year reservation, billed every hour whether the VM runs or not.

IONOS Cloud: IONOS bills Cubes by the minute at a fixed rate per size, with the NVMe disk included, and vCPU servers per core and per GB of RAM, with no plan fee. A suspended Cube keeps billing its storage, so the low workday figure assumes you snapshot and delete the Cube each night. Reserved IPv4 addresses (5 EUR per 30 days), extra storage and traffic past 2 TB a month cost extra.

IONOS Cloud: Plans cost $20, $100, $500 or $2,000 a month, and the whole plan price comes back as VM time for that month. A default 4 vCPU / 8 GB sandbox costs $0.036 an hour, billed per second while it runs. The $20 plan includes 12 starts a minute, 60 an hour and 200 a day; extra starts cost $0.03 to $0.07 each. A stopped sandbox has no compute charge. Saved snapshots above the plan's count and the AI credentials proxy still bill while stopped.

Limits in those scenarios

Azure caps vCPUs per region and per VM family, stopped VMs still count, and new subscriptions start low until you ask for more. IONOS does not publish how many machines an account can run at once, so 300 sandboxes for your users and 1,000 short jobs at once are not confirmed there. For the 1,000-job burst, the Azure figure uses burstable B-series VMs that guarantee 20% to 40% of each vCPU depending on size. On the $2,000 plan, boat charges for starts above 90 a minute, refuses them above 270 with 429, and never queues a start. One IONOS Cube tops out at 16 vCPU and 64 GB, a vCPU server at 60 vCPU and 230 GB, and the largest boat size at 16 vCPU and 32 GB.

ScenarioAzure Virtual MachinesIONOS Cloudboat
One coding agent, workdaysNone for this scenarioNone for this scenarioNone for this scenario
One agent that never stops
  • 3-year commitment
None for this scenarioNone for this scenario
50 agents in parallelNone for this scenario
  • how many machines can run at once: limit not published, probably fine at 50 (the typical published limit is 50)
None for this scenario
Sandboxes for your usersNone for this scenario
  • not confirmed: how many machines can run at once (you need 300; the typical published limit is 50)
None for this scenario
1,000 short jobs at once
  • burstable CPU: 30% of each vCPU guaranteed, throttled above that
  • not confirmed: how many machines can run at once (you need 1,000; the typical published limit is 50)
  • 2,400 starts/day included, paid above that (needs ~6,667)
  • 600 starts/hour included, paid above that (needs ~1,000)
  • starts above 90/min are paid, and starts above 270/min fail with 429

Which one should you use?

Use Azure Virtual Machines when

you need a region outside Europe and the US, Arm or burstable sizes, hibernation, Spot pricing, domain-based outbound rules, or your company already runs on Azure.

  • Regions and sizes almost everywhere. Useful if you need a region outside Europe and the US, or Arm, burstable and GPU sizes in one place.
  • Hibernation and Spot. Useful if a machine must resume with its memory, or a big batch can survive eviction with 30 seconds of notice.
  • Domain allowlists for outbound traffic. Useful if agents may only reach a fixed list of sites: Azure Firewall can limit outbound HTTPS by domain.

Use IONOS Cloud when

list price per machine-hour matters most, or when you want plain VMs from a German provider, a static IPv4, lots of free outgoing traffic, or H200 GPUs in Frankfurt and Worcester.

  • Low list prices on per-minute servers. Useful if price per machine-hour matters most: a Basic Cube M (4 vCPU / 8 GB) costs about $0.027 an hour with a 240 GB NVMe disk and no plan fee.
  • 2 TB of free outgoing traffic. Useful if your machines send a lot of data out: the first 2 TB each month cost nothing, against 100 GB on Azure.
  • A German company with H200 GPUs. Useful if you want a German provider with servers in Europe and the US, and H200 GPU VMs in Frankfurt and Worcester.

Where boat fits

Azure and IONOS both rent empty VMs, which is right for servers, GPUs and Windows. If you run agents, each one needs the same layer built on top: agents driven from outside, a desktop, keys kept out, snapshots and forks. boat gives every sandbox that layer on a full VM in Germany, Finland or France. IONOS has lower list prices in every scenario we priced.

Coding agents you drive from outside

On Azure or IONOS you install each agent, put API keys into the VM and write the code that prompts and watches it. boat ships the agents with a prompt, steer and events API. Its optional paid proxy keeps the real keys of supported AI providers out of the sandbox.

A desktop for computer use

On both clouds you install a desktop and an RDP, VNC or WebRTC server yourself. Every boat sandbox streams a 60 fps desktop with Chrome and computer-use tools.

Snapshots every minute, forks of a running machine

On Azure a copy starts from a gallery image or a disk snapshot and boots cold. On IONOS a snapshot is a manual full copy, and a suspended Cube keeps billing its storage. boat snapshots every minute and forks a running sandbox in one call; the fork can miss the last minute of writes.

Use boat when the machines are for AI agents: full VMs with Docker and /dev/kvm, coding agents you drive from outside, a desktop, snapshots every minute, forks and no compute charge while stopped. Compare it one to one: boat vs Azure Virtual Machines and boat vs IONOS Cloud.

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 Azure Virtual Machines cheaper than IONOS Cloud?

No. IONOS's list prices are lower in every scenario we priced. One agent running all month is $20 on IONOS and $48 on Azure (with a 3-year reservation), and 50 agents in parallel come to $240 against $1,168. boat comes to $26 and $317 for the same two. IONOS's workday figure of $5 assumes you delete the Cube each night, since a suspended Cube keeps billing its storage.

Which is faster, Azure or IONOS?

Neither is in the public benchmarks we use, so we cannot say. Their CPUs depend on the size or location you pick. boat did 28.06 V8 Web Tooling Benchmark runs per second at 4 vCPU / 8 GB in StarSling's run of 3 October 2026, on an AMD Ryzen 9 9950X host.

Can Azure and IONOS VMs run Docker?

Yes, both give you a root Linux VM where Docker runs. Azure supports nested virtualization on many Intel and AMD sizes, such as D, E and F series. IONOS documents nested Windows VBS and WSL 2, but has no guide for KVM in a Linux guest. boat has Docker, systemd and /dev/kvm ready on every sandbox.

Do Azure and IONOS keep billing when a machine is stopped?

Partly. Azure stops compute billing only after a deallocate, and disks and static IPs keep billing. On IONOS, a suspended Cube or a stopped vCPU server stops compute billing, but its storage keeps billing; to pay nothing you snapshot and delete it. A stopped boat sandbox has no compute charge, and boat keeps its disk for at least 14 days. Saved snapshots above the plan's count and the AI credentials proxy still bill.

Which has more regions?

Azure lists 77 regions worldwide. IONOS has data centers in Germany, the UK, France, Spain and the US, and you pick one. boat sandboxes run on machines in Germany, Finland and France, and you cannot pick a region. boat can store snapshot copies outside the EU.

Can I use Azure or IONOS as sandboxes for AI agents?

Yes. Both give each agent a full VM with its own kernel, which isolates code well. You build the sandbox layer yourself: agents driven by API, a desktop stream, a keys proxy, frequent snapshots and forks. On boat those come with every sandbox.

Try boat

Install the CLI, then run boat new.

7-day free trial

More comparisons