Uptime SLA

This service level agreement (SLA) describes Compute Cheap's uptime commitment for reserved GPU capacity, our support response targets, and the credits available if either is not met. It applies to prepaid reserved capacity blocks. It does not apply to interruptible capacity, which is preemptible by design and carries no uptime commitment.

This SLA is part of the agreement between you and Compute Cheap for your reserved capacity request. Capitalized terms not defined here have the meaning given to them in your capacity request confirmation.

Compute Cheap will use commercially reasonable efforts to make reserved capacity available with 99.9% monthly uptime, measured over each calendar month for each reserved GPU-hour block.

Monthly uptime percentage is calculated as: (total minutes in the month − downtime minutes) / total minutes in the month × 100. Downtime means the reserved GPUs are unreachable or fail to execute submitted workloads due to a cause within Compute Cheap's control.

If actual monthly uptime for a reserved block falls below 99.9%, you may request a service credit against that block's next invoice, calculated as a percentage of the fees paid for that block in the affected month:

Service credits are your sole and exclusive remedy for Compute Cheap's failure to meet the uptime commitment in this SLA. Total credits for any month cannot exceed 50% of that month's fees for the affected block.

To request a service credit, email hello@compute.cheap within 30 days of the end of the affected month, including the reservation ID and the dates and times of the downtime you observed. We'll confirm eligibility against our monitoring records and apply approved credits to your next invoice.

Reserved capacity is monitored 24 hours a day, every day of the year. We target the following first-response times once an issue is reported or detected:

Report an issue by emailing hello@compute.cheap. Include your reservation ID so we can respond as quickly as possible.

The uptime commitment in this SLA does not apply to downtime caused by:

• Interruptible capacity, which is preemptible by design.
• Scheduled maintenance, announced at least 48 hours in advance.
• Factors outside Compute Cheap's reasonable control, including upstream network or power outages, and force majeure events.
• Your equipment, software, workloads, or network configuration.
• Suspension of your account for breach of the acceptable use policy or nonpayment.

We may update this SLA from time to time. If a change materially reduces the uptime commitment or support targets for capacity you have already reserved, we'll give you at least 30 days' notice by email before it takes effect for that reservation.