Service Level Agreement

Enterprise agreements target 99.95% monthly availability of the gateway API, measured at our edge. Provider outages are not gateway downtime; that's what cross-provider failover is for, and effective success is reported separately on the status page.

If monthly availability falls below the target, you're entitled to service credits against our fees: 10% below 99.95%, 25% below 99.5%, 50% below 99.0%. Claims go to your named engineer within 30 days of the incident; credits are the exclusive remedy for availability misses.

P1 (production down): response within 4 hours, 24/7, in your shared Slack channel. P2 (degraded): next business day. P3 (questions): two business days. All targets are placeholder values pending the console launch.

Downtime caused by your infrastructure or tunnel misconfiguration, provider-side outages of specific models, force majeure, or scheduled maintenance announced 72 hours ahead (max 4 hours monthly, off-peak Eastern) doesn't count against the target.