AI Latency Tracker

Latency Tracker monitors the response speed and availability of inference APIs from major providers across several global regions. Key features: • Latency probes from Germany, US Central, Tokyo and São Paulo • Probes run every 5 minutes against official API endpoints • Real uptime history and outage reports • Model retirement and deprecation dates • Free JSON API and MCP access The tracker shows which provider responds fastest from your location, so you can choose endpoints that match your latency needs. Independent and provider-neutral, it does not favor any vendor. Historical charts reveal reliability trends, while incident logs help you understand when and why disruptions happened. Teams can compare time to first token, pricing against speed, and uptime across 44 providers. Data is published under CC BY 4.0 and can be pulled into dashboards, scripts or coding agents through the free API. Ideal for developers, platform engineers and product teams who depend on reliable model access.