Machine Catalogs
Varity publishes four separate machine catalogs. Each one is a different product with a different price shape and a different way of getting a machine. Pick the catalog first.
Every number on this page was read from GET /api/deployment-profiles on 22 August 2026.
The Four Catalogs
Section titled “The Four Catalogs”| Catalog | workload | execution_class | Profiles | Regions | Hourly price |
|---|---|---|---|---|---|
| CPU VMs | virtual_machine | cpu_virtual_machine | 26 | 2 | $0.019178 to $0.89726 |
| GPU containers | gpu | container | 9 | not returned | $1.431 to $5.13 |
| GPU VMs | gpu | virtual_machine | 186 | 53 | $0.513 to $59.562 |
| GPU bare metal | gpu | bare_metal | 4 | 4 | $53.4735 |
Both query parameters are required. Every profile in all four catalogs was available at the time of reading.
What You Get
Section titled “What You Get”| Catalog | You rent | How you create it | Access | Monthly cap |
|---|---|---|---|---|
| CPU VMs | A CPU virtual machine, 2 to 64 vCPU | POST /api/machines | SSH | Yes |
| GPU containers | A container running on one GPU | POST /api/deployments | Container deployment, no SSH | No |
| GPU VMs | A VM with 1, 2, 4, or 8 attached GPUs | POST /api/machines | SSH | No |
| GPU bare metal | A physical 8-GPU host, no hypervisor | POST /api/machines | SSH | No |
Only CPU VMs carry monthly_cap_microusd. GPU profiles do not have a cap at all. A forgotten GPU machine accrues indefinitely.
Which One
Section titled “Which One”| You want to… | Use |
|---|---|
| Run a worker, daemon, database, or build agent you administer | CPU VMs |
| Control the kernel or install system packages, cheaply | CPU VMs, from $0.019178/hour |
| Run one container on one GPU without managing a host | GPU containers, 4 card types, $1.431 to $5.13/hour |
| Train or fine-tune on a specific card, with SSH | GPU VMs, 21 card models, 53 regions |
| Get a whole 8× RTXPRO6000 host with no virtualization layer | GPU bare metal, $53.4735/hour |
| Ship a web app and get a URL, with no machine to maintain | Applications |
| Call a hosted model instead of running one | AI Gateway |
GPU Containers Use a Different Schema
Section titled “GPU Containers Use a Different Schema”CPU VM, GPU VM, and GPU bare-metal profiles share one shape (schema_version: public-machine-profile-v1). GPU container profiles are deliberately different (gpu-container-profiles-v1). Do not write one parser for all four.
| Machine profiles | GPU container profiles | |
|---|---|---|
| Identifier | profile_id, prefix mp- | id, prefix gpo- |
| Human name | display_name | label |
| Price field | customer_price.amount_microusd (integer microdollars) | customer_price.hourly_usd (decimal dollars) |
| Host hardware | cpu_units, memory_mb, storage_mb | none (card only) |
| GPU facts | hardware.accelerator.alternatives[] | hardware and accelerator.alternatives[] |
region | present | not returned |
os_images | present | not returned |
availability | a string, e.g. "available" | an object: status, matching_capacity_count, aggregate_available_units, source_age_seconds |
| Created via | POST /api/machines | POST /api/deployments |
Reading a Catalog
Section titled “Reading a Catalog”curl "https://varity.app/api/deployment-profiles?workload=virtual_machine&execution_class=cpu_virtual_machine" \ -H "Authorization: Bearer $VARITY_API_KEY"curl "https://varity.app/api/deployment-profiles?workload=gpu&execution_class=virtual_machine" \ -H "Authorization: Bearer $VARITY_API_KEY"Swap execution_class for container or bare_metal to read the other two GPU catalogs.
Prices
Section titled “Prices”Every catalog quotes billing_period: "hour" in USD.
- CPU VMs publish
amount_microusdandmonthly_cap_microusd. The hourly rate is the monthly cap divided by 730. - GPU VMs and bare metal publish
amount_microusdonly. - GPU containers publish
hourly_usddirectly, withprice_basis: "exact"andpricing_policy_version: "gpu-hourly-v2-2026-07-28".
Catalog prices are the list rate at the moment the catalog was read. CPU VM rates are fixed: the hourly figure is the monthly cap divided by 730. GPU rates are not a fixed price list. They follow the capacity actually on offer, so the GPU figures on this page are a dated snapshot rather than a rate card. See GPU machines for the detail.
Before creating anything you request a quote, which returns the binding hourly_usd, an authorization_usd, and a valid_until timestamp.