AI reliability
AI assistants, model APIs, and chat platforms — ChatGPT, Claude, Gemini, and others.
By estimated reliability
Sorted by reliability score, descending. Each row links to the service profile.
What stands out
What we’re tracking
This chart aggregates reliability signals for ai services: external probes that run every five minutes, community outage reports filed when real users hit a problem, and the service’s own status-page declarations where one is available.
The reliability score is an estimate from those signals — not a direct uptime measurement. Use it to compare services against each other and to spot trend changes, then click through to a service profile for the detailed history. Reports counts are raw community signal and tend to track real disruption with some delay.
See the methodology page for the exact thresholds and scoring inputs.
WebsiteDown runs probes, alerts, dashboards, and incident tooling — point it at the ai services you depend on and get the same signal we surface here.