New B200 spot capacity is live in US East from $1.69 per GPU-hour. See availability →
Deploy a GPU
Service status

All systems operational.

Checked when this page was rendered, against the same database and capacity pool the console uses. No cached green dots.

Rendered 16 September 2026, 09:11 UTC

Components
5All reporting normally
Capacity now
2,392GPUsFree for a new instance across 3 regions
Uptime target
99.9%On-demand instances, monthly — see the SLA
Since
14 Sep 2026Platform launch; this record starts there
Components

What each part is doing right now.

Platform changelog
Public websitespotgpus.com, price pages and documentation.Served from the edge; this page was rendered by the origin.
Operational
Customer consolecloud.spotgpus.com — accounts, deployment, billing and support.Session store and account database reachable.
Operational
Catalogue and pricingGPU models, node sizes and the prices shown across the site.Last price refresh 24 hours ago.
Operational
Capacity poolGPUs advertised as available for a new deployment.2,392 GPUs available across 3 regions.
Operational
PaymentsCredit top-ups and payment confirmation.Deposits are confirmed on-chain; a payment already sent is credited even if this component is interrupted.
Operational
Capacity

GPUs free for a new instance, by region.

These are the numbers the deployment screen books against. They move through the day as jobs start and end; a zero means the pool is empty right now, not that the region is down.

US EastVirginia, United States 528
21 models offered Tier III us-east
US WestOregon, United States 928
34 models offered Tier III us-west
EU CentralFrankfurt, Germany 936
33 models offered Tier III eu-central

Per-model availability is on the catalogue and on each model page. Need capacity guaranteed on a date? Reserve it.

Incidents

The record since launch.

Incident history

since 14 September 2026

No customer-affecting incident has been recorded. The platform opened in private beta on 14 September 2026; that is a short record, and we would rather say so than dress it up as a track record.

When something does break, it will appear here with the date, what was affected, how long it lasted and what we changed afterwards — including the ones that were our fault.

A spot reclaim is not an incident.

Reclaiming spot capacity with a 2-minute notice is how the tier works, and it is never reported here. How spot works · the tier that is never reclaimed

What we commit to

On-demand uptime
99.9% monthlyWith service credits below it. The SLA defines the measurement and the exclusions.
Spot instances
No uptime targetThey are interruptible by design, with 2 minutes of notice.
First reply on a broken instance
Within 2 hoursTickets are answered by the engineers who operate the nodes, 24/7.
Notice of maintenance
Announced in advancePlanned work that can affect running instances is announced to account holders by email and shown here.

Open a ticket

Get started

Nothing to report is the point.

Capacity is live, the catalogue is fresh and the console is up. Add credit and run something.

Deploy a GPU See the prices

Billed per minute from the moment the instance is reachable. Minimum credit $40, no subscription.