SLA with real accountability
We accept financial accountability for service availability. Public status page, SLA credits on breach, 15-minute P1 response, and post-mortems within 5 business days.
Our commitments
Contractual SLOs codified in the SLA appendix to your master agreement.
99.9% uptime
Monthly API availability no lower than 99.9%. On breach, SLA credits are applied automatically to the next billing cycle.
P1 in 15 minutes
Named on-call SRE responds to critical incidents within 15 minutes of alert, 24/7, no exceptions.
Auto-failover in 30s
Health polling on every upstream every 30 seconds. Automatic reroute to a fallback path on degradation.
Post-mortem in 5 days
Public post-mortem for every P1 incident within 5 business days: root cause, timeline, remediation plan.
SLA credits
When we miss the SLO, you get a portion of your monthly spend back as a balance credit.
| Monthly availability | SLA credit |
|---|---|
| < 99.9% and ≥ 99.0% | 10% of monthly usage |
| < 99.0% and ≥ 95.0% | 25% of monthly usage |
| < 95.0% | 50% of monthly usage |
Credits are applied automatically at month-end. No ticket needed — the credit shows up in the dashboard and offsets subsequent requests.
Exclusions
The SLO does not cover events outside our control.
- Planned maintenance announced at least 72 hours in advance.
- Force majeure (natural disasters, government action).
- Degradation at an upstream model provider while our fallback route is active.
- Customer-caused issues: DDoS of your own keys, rate-limit violations, integration errors.
- Network problems between the customer's edge and our perimeter (BGP, ISP, CDN on customer side).
Request the full SLA
The SLA document is part of the enterprise MSA and DPA. Terms are tailored to customer-specific requirements.