System status.
Live availability across 14 regions and 8 services. Rolling 12-month uptime is 99.98%.
All systems operational
Services
90-day uptime
Regions
Method
How to read this page
How uptime is measured
Every figure here comes from external probes hitting each region once a minute, not from our own health checks. A region counts as down if provisioning fails or if the streaming gateway stops accepting sessions — a device that keeps running while the console is unreachable still counts as an outage, because you cannot operate what you cannot see.
What the SLA covers
We target 99.9% monthly availability for provisioning and streaming on every plan. Enterprise contracts can buy a contractual SLA at 99.95% with service credits and a 30-minute response commitment. Scheduled maintenance announced at least 48 hours ahead is excluded, and we have used that window four times in six years.
What we publish
Every incident, including the ones that cost us money. The November 2025 entry below is a 61-hour failure that lost 1,847 customer accounts and roughly $91,000 in credits and engineering time. It is here for the same reason the ban rates are published: numbers you can only see when they are flattering are not numbers, they are marketing.
Recent incidents
Elevated streaming latency in br-south
A transit provider in São Paulo degraded, adding 180–400 ms to device streaming for customers in that region. Devices continued running normally; only the interactive view was affected. Rerouted at 14:41 UTC.
API rate limiter misconfiguration
A deploy applied Starter rate limits to a subset of Growth accounts, returning 429s on legitimate calls. Rolled back within 18 minutes. No device impact.
Snapchat Device Attestation v3
Snap shipped a substantially reworked keymaster check without warning. Our patch took 61 hours. 1,847 customer accounts were lost. All were credited under Ban Shield and rebuilt free of charge. Full post-mortem published in the changelog.