Service Level Agreement

The Short Version

We guarantee 99.9% uptime for your GPU instances. If we fail to meet this, you get credits back proportional to the downtime. This page explains exactly how that works, what counts as downtime, and what doesn't.

1. Our Uptime Commitment

We commit to 99.9% monthly uptime for all GPU instances. This means your instances should be available at least 99.9% of the time in any given calendar month.

In practical terms, 99.9% uptime allows for up to 43 minutes and 50 seconds of downtime per month.

2. Service Credits

If we don't meet our 99.9% uptime commitment, you're entitled to service credits based on the following schedule:

Credits are applied to your account balance for future use. The maximum credit for any month is 50% of your charges for that month. Credits don't expire but can't be converted to cash.

3. What Counts as Downtime

Downtime means your GPU instance is completely unavailable and you cannot:

• SSH into your instance
• Access your running workloads
• Reach services running on your instance

We measure downtime from when we confirm an issue (either through our monitoring or your support ticket) until service is restored.

4. What Doesn't Count as Downtime

The following situations are excluded from our uptime calculation:

• Scheduled maintenance: We'll notify you at least 48 hours in advance for planned maintenance windows.
• Your code or configuration: Issues caused by your applications, scripts, or misconfiguration.
• Resource exhaustion: Running out of disk space, memory, or hitting compute limits you configured.
• Network issues outside our control: Problems with your ISP, DNS providers, or internet backbone issues.
• Abuse or violations: Service suspension due to terms of service violations or abusive behavior.
• Force majeure events: See section 6 below.

5. How to Request Credits

To request SLA credits:

1. Email us at support@packet.ai within 30 days of the incident.
2. Include: Your account email, affected instance IDs, date and time of the incident, and a brief description.
3. We'll review your request within 5 business days and respond with our determination.
4. If approved, credits will be added to your account balance.

6. Force Majeure

We're not responsible for failures caused by circumstances beyond our reasonable control. This includes, but isn't limited to:

• Natural disasters (earthquakes, floods, hurricanes, etc.)
• Wars, terrorism, or civil unrest
• Government actions or court orders
• Widespread power grid failures
• Major internet infrastructure outages affecting multiple providers
• Pandemics or public health emergencies
• Acts of God

During force majeure events, we'll do our best to restore service as quickly as possible and keep you informed of our progress.

7. Our Responsibilities

We commit to:

• Monitoring our infrastructure 24/7
• Responding to critical issues within 1 hour
• Providing status updates during incidents via our status page
• Conducting post-incident reviews for significant outages
• Continuously improving our infrastructure reliability

8. Your Responsibilities

To benefit from this SLA, you should:

• Keep your contact information up to date
• Report issues promptly through our support channels
• Implement reasonable redundancy for critical workloads
• Back up your data regularly (we provide the infrastructure, you own your data)
• Stay within your resource limits and terms of service

9. Limitations

This SLA applies only to paid GPU instances. Free tiers, beta features, and promotional offerings may have different or no uptime guarantees.

Service credits are your sole and exclusive remedy for any failure to meet our uptime commitment. This SLA doesn't create any additional warranties or liabilities beyond what's stated here.

10. Changes to This SLA

We may update this SLA from time to time. If we make changes that reduce our commitments, we'll notify you at least 30 days in advance. Continued use of our service after changes take effect means you accept the updated terms.

Questions?

If you have questions about this SLA or need to report an incident, contact us at support@packet.ai.