Reference

machine0 Pricing: Per-Minute Cloud VMs and GPUs

Last updated: November 14, 2026

machine0 bills per minute, pay as you go, at the same price in all nine regions. CPU VMs run from $0.013/hr (small, 1 vCPU, 1 GB) to $3.714/hr (6xl, 60 vCPU, 240 GB). GPU VMs run from $0.836/hr (1x RTX 4000 Ada) to $39.336/hr (8x H200). Suspended VMs pay only image storage at $0.078/GB/month. The minimum wallet top-up is $5 and unused credits are refundable.

## CPU VMs

Prices per hour, billed per minute. The monthly column assumes 730 hours.

SizevCPURAMDisk$/hour~$/month
small11 GB25 GB$0.013$9
medium22 GB60 GB$0.034$25
large24 GB80 GB$0.052$38
xl48 GB160 GB$0.104$76
xxl816 GB320 GB$0.208$152
large-nvme24 GB80 GB$0.061$45
xl-nvme48 GB160 GB$0.121$88
xxl-nvme816 GB320 GB$0.243$177
xl-premium48 GB50 GB$0.236$172
xxl-premium816 GB100 GB$0.473$345
xxxl1664 GB200 GB$0.825$602
4xl32128 GB480 GB$1.980$1,445
5xl48192 GB720 GB$2.970$2,168
6xl60240 GB900 GB$3.714$2,711

The -nvme sizes use newer CPUs with NVMe storage for higher disk IOPS, with the same vCPU, RAM and disk as their regular counterparts. The -premium sizes use dedicated vCPUs, the fastest single-thread performance on offer.

A region is a specific datacenter, and your VM, its disks and its snapshots live exactly there. CPU sizes run in nyc3 (New York), sfo3 (San Francisco), lon1 (London), ams3 (Amsterdam) and sgp1 (Singapore). The old names us-east, us-west, uk, eu and asia still work and map to those datacenters.

## GPU VMs

Available in tor1 (Toronto), nyc2 (New York), atl1 (Atlanta), fra1 (Frankfurt), lon1 (London), ams3 (Amsterdam) and sgp1 (Singapore), not nyc3 or sfo3. Pick a GPU datacenter explicitly with --region; machine0 sizes shows what is in stock where.

SizeGPUVRAMvCPURAMDisk$/hour~$/month
gpu-4000ada-11x RTX 4000 Ada20 GB832 GB500 GB$0.836$610
gpu-l40s-11x L40S48 GB864 GB500 GB$1.727$1,261
gpu-6000ada-11x RTX 6000 Ada48 GB864 GB500 GB$1.727$1,261
gpu-mi300x-11x MI300X192 GB20240 GB720 GB$2.849$2,080
gpu-h100-11x H10080 GB20240 GB720 GB$4.851$3,541
gpu-h200-11x H200141 GB24240 GB720 GB$4.917$3,589
gpu-mi300x-88x MI300X1,536 GB1601,920 GB2 TB$22.792$16,638
gpu-h100-88x H100640 GB1601,920 GB2 TB$38.808$28,330
gpu-h200-88x H2001,128 GB1921,920 GB2 TB$39.336$28,715

If a GPU size is out of stock, machine0 order places a standing order that is created automatically when capacity appears and is not billed until the machine runs.

## What gets billed

ItemRateNotes
MachinesThe size's hourly rate, per minuteBilled while running and while stopped
Images and snapshotsAbout $0.078/GB/monthThis is what a suspended VM pays
DisksAbout $0.1667/GB/monthUntil removed; survive suspend and VM destruction
Bandwidth$0.01/GiBBeyond a free allowance of 1 to 10 TB depending on size

The practical rule: stop does not save money, suspend does. Suspend snapshots the disk, deletes the cloud instance, and leaves you paying image storage only. Everything on disk survives; running processes do not, and the VM gets a new IP on resume.

## Wallet and top-ups

machine0 is pay as you go from a prepaid wallet.

  • Minimum top-up $5, maximum $10,000 per top-up.
  • Unused credits are refundable.
  • Monthly PDF invoices.
  • Until your first top-up, you can run 1 machine. After it, 50 at once for accounts created since July 28, 2026 (older accounts keep their previous limit; contact support to raise it).

## The suspension threshold

When your balance reaches the suspension threshold, machine0 protects your data rather than your uptime: running machines are snapshotted and destroyed, and the snapshots are kept. The threshold is the amount needed to store your snapshots, images and disks for 7 days, usually a few dollars.

A warning email arrives about 6 hours before. machine0 account shows the threshold and the estimated time to suspension.

## Auto top-up

Auto top-up is on by default once a card is saved, and it can be turned off on the Account page. With it on, the wallet refills before it runs dry. With it off, your balance is a hard spending ceiling: nothing can be created once it is spent. There is no separate spend cap or budget alert feature, so this toggle is the control.

## Cost examples

ScenarioMathCost
A large VM around the clock for a month730 h x $0.052About $38
A large VM 8 hours a day for 22 workdays, suspended in between176 h x $0.052About $9.15
One H100 for 3 hours3 h x $4.851About $14.55
8x H200 for 1 hour1 h x $39.336$39.34
OpenClaw on medium 24/7730 h x $0.034About $25

Add image storage for any time a VM spends suspended: an 80 GB large disk is about $6.24/month at $0.078/GB.

## Getting started

$ curl -LsSf https://machine0.io/install.sh | sh
$ machine0 new my-vm --size large
$ machine0 ssh my-vm

## FAQ

Is a stopped VM free?

No. A stopped VM bills at the full rate. Suspend it instead: the disk is snapshotted, the instance is deleted, and you pay only image storage.

What is the suspension threshold?

The balance at which machine0 protects your data: enough to store your snapshots, images and disks for 7 days, usually a few dollars. At that point running machines are snapshotted and destroyed and the snapshots are kept. You get a warning email about 6 hours before.

Can I set a spending cap?

Turn auto top-up off on the Account page. Your wallet balance then becomes the most that can be spent. There is no separate spend cap or budget alert feature.

Are GPU prices the same in every region?

Yes. GPU sizes run in the GPU-capable datacenters (tor1, nyc2, atl1, fra1, lon1, ams3 and sgp1) at the same price. They are not offered in nyc3 or sfo3.

Do I pay for disks while a VM is suspended?

Yes. Persistent disks bill at about $0.1667/GB/month until you remove them, and they survive suspend and VM destruction. The VM's own disk is stored as an image at $0.078/GB/month while suspended.

Can I get a refund?

Unused credits are refundable. Top-ups are between $5 and $10,000.

## Sources

## Related

  • GPU VMs billed by the minute: H100, H200, MI300X, L40S and RTX Ada sizes, the NVIDIA image, standing orders and suspend.
  • Cloud computers for AI agents: Persistent VMs vs sandboxes, with a comparison of machine0, E2B, Daytona and Fly.io Sprites.
  • Rate limits: API, machine, wallet and MCP limits, and how to handle 429 responses.
  • About: What machine0 is, who builds it, and how to get in touch.