Nebius vs Jarvislabs
Both rent Nvidia GPUs and plain CPU virtual machines, but Nebius is a full cloud with Kubernetes, Slurm, big machines and regions in the EU, UK, Israel and US. Jarvislabs is a smaller GPU cloud that bills by the minute from a prepaid wallet and creates new machines only in India.
The companies and products
Nebius
Nebius is an AI cloud company, listed on Nasdaq and based in Amsterdam, that rents GPUs and the infrastructure around them.
You create a Compute VM from a preset (GPU or CPU only), add disks and an SSH key, and get a full Linux VM with root and an optional public IP. Around it Nebius sells Managed Kubernetes, Slurm clusters (Soperator), Serverless AI jobs and endpoints, object storage and networking. It is built for teams that train, fine-tune and serve AI models, not as a sandbox for agents.
Jarvislabs
Jarvislabs is an AI cloud from India that rents Nvidia GPUs by the minute for training and serving models.
You can rent a GPU VM with root and your own kernel, a Template (a GPU container with PyTorch, ComfyUI or similar ready, opened in JupyterLab, VS Code or SSH), a plain CPU VM, or a multi-node GPU cluster. It also serves open models through Managed Endpoints, Serverless workers and a token-priced Model API. It is built for people who train, fine-tune and serve AI models, not as a sandbox for agents, though its jl CLI is written so a coding agent on your laptop can drive the GPUs.
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. “29th cheapest of 364” means 28 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 | Nebius: price rank | Jarvislabs: price rank | bochka: price rank |
|---|---|---|---|
| Software factory 100 at once · 4 vCPU / 8 GB · 2,200 sessions of 8 h a month Neither ships a coding agent, so on both you install the agent, Docker and tooling yourself, while bochka ranks 3rd with agents ready inside every sandbox. | 197th of 367, with limits $2,272 a month | 154th of 367, with limits $1,406 a month | 3rd cheapest of 367 $634 a month |
| Always-on AI employees 50 machines 24/7 · 2 vCPU / 4 GB | 210th of 366, with limits $2,261 a month | 185th of 366, with limits $1,670 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 bursts suit the two pay-as-you-go clouds better than bochka, whose start limits make it 186th here. | 111th of 345, with limits $202 a month | 103rd of 345, with limits $169 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 Nebius ranks 81st and Jarvislabs 215th: Nebius starts at 12 VMs per region and raises that on request, while Jarvislabs does not publish its limit, so 300 user sandboxes means asking either one first. | 81st of 361, with limits $716 a month | 215th of 361, with limits $539 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 | 52nd of 360, with limits $3,049 a month | 43rd of 360, with limits $2,568 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 Nebius is a full match for agent evals at 29th, Jarvislabs a compromise at 113th, and bochka ranks 10th with snapshots every minute and a fork of a running sandbox. | 29th cheapest of 364 $610 a month | 113th of 364, with limits $411 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 All three fit as CI runners; bochka ranks 4th, with per-second billing and stops that cost nothing between jobs. | 27th cheapest of 377 $297 a month | 20th cheapest of 377 $253 a month | 4th cheapest of 377 $86 a month |
| 24/7 dev boxes 1 machine 24/7 · 4 vCPU / 8 GB | 68th cheapest of 368 $91 a month | 52nd cheapest of 368 $68 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 Nebius nor Jarvislabs has a desktop or computer-use product, while bochka streams a 60fps desktop with Chromium and ranks 6th. | 187th of 369, with limits $610 a month | 154th of 369, with limits $377 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 | 223rd of 367, with limits $305 a month | 204th of 367, with limits $205 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 | 253rd of 346, with limits $302 a month | 244th of 346, with limits $247 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 | 174th of 338, with limits $6,049 a month | 158th of 338, with limits $4,942 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 | 176th of 360, with limits $156 a month | 156th of 360, with limits $127 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, and Jarvislabs ranks 9th at $876 a month against Nebius at 18th and $1,127; bochka has no GPUs. | 18th cheapest of 98 $1,127 a month | 9th cheapest of 98 $876 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 | Nebius | Jarvislabs | bochka |
|---|---|---|---|
| Kind of product | AI cloud: GPU and CPU VMs, Kubernetes, Slurm clusters, serverless jobs and storage | GPU cloud: GPU VMs, GPU containers, CPU VMs, clusters and model serving | Sandboxes for AI agents: VMs with agents, desktop and hosting ready |
| Isolation | Every Compute VM is a full VM behind a hypervisor | GPU and CPU VMs are KVM virtual machines; Templates are containers | Every sandbox is a full VM with its own kernel |
| GPUs | Nvidia L40S, RTX PRO 6000, H100, H200, B200 and B300, on-demand or preemptible | Nvidia L4 up to H200, 1 to 8 per instance, on-demand, spot or reserved; B200 as reserved clusters | None |
| Biggest CPU machine | Up to 256 vCPU / 1,024 GB | Up to 32 vCPU / 128 GB | 16 vCPU / 32 GB on the $100 plan and up |
| Docker and nested virtualization | Root on your own VM and container VMs from Docker images; nested virtualization not documented | Root on VMs, so you can run your own Docker daemon; nested virtualization not documented | Docker, root, systemd and /dev/kvm on every machine |
| Billing | Per second, pay as you go, no plan fee | Per minute from a prepaid wallet, $10 minimum top-up, no free trial | Monthly plan from $20, all of it usage credit spent per second |
| Stopping a machine | Compute stops billing; disks keep billing on their full size | Pause stops compute billing; the disk keeps billing, and VMs need at least 100 GB | A stopped sandbox costs nothing and keeps its disk |
| Snapshots and fork | Manual disk snapshots of a running VM; fork by making a disk, then a VM, from the snapshot | No snapshot, clone or fork command | Snapshots every minute and on stop; fork a running sandbox in one call |
| Desktop and computer use | No desktop or browser product; a text serial console | No desktop; JupyterLab, VS Code in the browser and SSH | 60fps desktop and Chromium, computer use and browser control, embeddable stream |
| Coding agents | None built in; you install agents yourself over SSH | A CLI and skill file so an agent on your own machine can manage GPUs | Claude Code, Codex, OpenCode and more preinstalled, with a prompt, steer and events API |
| Keeping keys out of the machine | SecretStash stores secrets that a VM fetches with its service account | Keys passed in as environment variables or uploaded files | AI credentials proxy for model providers and any HTTPS API you register |
| Network rules | VPC security groups for ingress and egress by IP range, port and protocol | Security groups with inbound and outbound rules, VPCs and reserved IPs | Open internet on every plan; no provider-side allowlist |
| Model serving that scales to zero | Serverless AI jobs and endpoints on Compute VMs | Serverless model workers stop after an idle timeout and start on the next request | No wake on request; you start a stopped sandbox through the API |
| Regions | Finland, France, Spain, United Kingdom, Israel and the US | India only for new instances (IN1, IN2) | EU only: Germany, Finland, France |
Every feature Battleships tracks (43)
| Feature | Nebius | Jarvislabs | bochka |
|---|---|---|---|
| Full VM (own kernel) | Yes | Yes | Yes |
| Docker inside | Yes | Yes | Yes |
| Nested virtualization | Not published | Not published | Yes |
| Root | Yes | Yes | 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) | Yes | 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 | Yes | No |
| HTTPS preview URLs | Yes | Yes | Yes |
| Raw TCP inbound | Not published | 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 | 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 | Not published | Yes | Yes |
| Usage per sandbox (API) | Yes | Yes | Yes |
| Audit logs | Not published | Not published | No |
| SSO | Yes | No | No |
| SOC 2 | Yes | No | No |
| HIPAA | Yes | No | No |
| EU data residency | Yes | No | Yes |
| Zero data retention | Not published | Not published | Yes |
Performance
There is no comparable CPU measurement for Nebius or Jarvislabs: neither was part of the public StarSling benchmark run. 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 hosts. Nebius CPU VMs run on AMD EPYC Genoa or Intel Ice Lake, and Jarvislabs publishes GPU benchmarks but none for its CPU VMs.
Node.js build-and-test loop at 4 vCPU / 8 GB, higher is better. Measured with the open hpc-sandbox-benchmarks suite. Nebius and Jarvislabs 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 | Nebius, a month | Jarvislabs, 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 | $25 · Pay as you go | $21 · On-demand | $20 · $20/mo |
| One agent that never stops 4 vCPU / 8 GB, 50 GB disk, running 24/7 all month | $91 · Pay as you go | $71 · On-demand | $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 | $1,127 · Pay as you go | $892 · On-demand | $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 | $305 · Pay as you go | $231 · On-demand · 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 | $610 · Pay as you go | $514 · On-demand · with limits | $2,000 · $2000/mo · with limits |
Nebius: Nebius bills running VMs per second for vCPU and RAM, with no plan fee, and bills disks on their provisioned size even while the VM is stopped. From 1 October 2026 an AMD Genoa CPU VM costs $0.015 per vCPU-hour plus $0.0045 per GiB-hour.
Jarvislabs: Jarvislabs bills per minute from a prepaid wallet with no plan fee. CPU VMs cost $0.012 per vCPU-hour plus $0.0032 per GB of RAM, and the disk is extra at $0.10/GB-month, with at least 100 GB on a VM, billed while running and paused.
Jarvislabs: 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
Nebius starts each region at 12 VMs and, in most regions, 200 vCPU across CPU-only VMs, raised on request, so 50 agents, 300 user sandboxes and 1,000 short jobs all need a bigger quota first. Jarvislabs does not publish how many machines an account can run at once, so 300 sandboxes or 1,000 jobs at once are not confirmed until you ask. Both start CPU VMs at 2 vCPU / 8 GB, and a Jarvislabs VM takes about 90 seconds to boot. On bochka the 1,000-job burst is queued and needs more starts per day than the $2,000 plan allows.
| Scenario | Nebius | Jarvislabs | 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 |
| 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 Nebius when
you need the newest GPUs, Slurm or Kubernetes clusters, a very large machine, or regions in the EU, UK, Israel or the US.
- The newest GPUs and full clusters. Useful for large training runs: Nebius rents B200 and B300 GPUs on demand and runs Slurm and Kubernetes clusters.
- Very large machines. Useful for one huge CPU job: Nebius CPU VMs reach 256 vCPU and 1,024 GB of RAM.
- Regions in the EU, UK, Israel and US. Useful when your data or users sit outside India: Nebius runs regions across Europe, Israel and the US.
Use Jarvislabs when
you want GPUs by the minute at a lower bill, model serving that scales to zero, or machines in India.
- Lower bills in every priced scenario. Useful on a tight budget: Jarvislabs came out cheaper than Nebius in every scenario we priced, from one agent to GPU evals.
- Model serving that scales to zero. Useful for inference: Jarvislabs serverless workers stop when idle and start again on the next request.
- GPUs by the minute in India. Useful for teams in India: machines run in Noida and Indian customers can pay in rupees.
Where bochka fits
Nebius and Jarvislabs are GPU clouds, so pick between them for training and serving models. If what you are really running is agents on CPU machines, bochka gives each one a ready VM with coding agents, a desktop, snapshots every minute and free stops. Many teams keep GPUs on one of these clouds and run the agents on bochka.
Agents and a desktop come ready
On Nebius and Jarvislabs a VM is a clean Linux machine, so you install the agent, a browser and any desktop, then write the runner that prompts and watches it. bochka ships Claude Code, Codex and a 60fps desktop on every sandbox.
Roll back and branch agent work
Jarvislabs has no snapshots or forks, and Nebius snapshots only by hand, with a fork costing a new disk and a new VM. bochka snapshots every minute and forks a running sandbox in one call.
Idle agents and real keys
A stopped Nebius VM still pays for its disks and a paused Jarvislabs VM bills at least 100 GB, while a stopped bochka sandbox costs nothing. Both put real API keys on the machine; bochka's credentials proxy keeps them outside.
Use bochka when you run agents on CPU machines: VMs that come with coding agents, a desktop, HTTPS ports and a credentials proxy, that snapshot every minute, fork while running and cost nothing while stopped. Compare it one to one: bochka vs Nebius and bochka vs Jarvislabs.
Questions people ask
Is Nebius cheaper than Jarvislabs?
No, Jarvislabs came out cheaper in every scenario we priced. One agent running all month is $91 on Nebius against $71 on Jarvislabs, and 50 agents in parallel are $1,127 against $892. On bochka those two come to $26 and $317 a month.
Which has better GPUs, Nebius or Jarvislabs?
Nebius rents B300 and B200 on demand plus H200, H100, RTX PRO 6000 and L40S. Jarvislabs rents L4 up to H200, 1 to 8 per instance, with B200 only as reserved clusters. For GPU eval jobs Jarvislabs comes to $876 a month against $1,127 on Nebius.
Which is faster?
We have no comparable CPU score for either: neither was in the public StarSling benchmark run. bochka did 26.69 Node.js build-and-test runs per second at 4 vCPU / 8 GB in that run.
Can Nebius and Jarvislabs run Docker?
Yes. Both give you root on a full VM, so you can run your own Docker daemon, and Nebius can also start container VMs from Docker images. Neither documents nested virtualization, while bochka has Docker, systemd and /dev/kvm on every machine.
Does a stopped machine cost money?
On both, the disk keeps billing. A stopped Nebius VM pays for its disks on their full size, and a paused Jarvislabs VM pays for at least 100 GB. A stopped bochka sandbox costs nothing and keeps its disk.
Where do Nebius and Jarvislabs run?
Nebius runs in Finland, France, Spain, the United Kingdom, Israel and the US. Jarvislabs creates new instances only in India; its old EU1 region no longer accepts new instances. bochka runs only in the EU, in Germany, Finland and France.
Try bochka
Install the CLI, then run bochka new.
Product preview