Stop finding out from your users that the AI is down. Get a real-time command-center view of every major AI API — straight from official feeds.
Open Live Dashboard →When an AI API goes down, most developers find out in the worst possible way — from a support ticket, a Slack message, or a spike in errors. By then, users are already frustrated.
Checking each provider's status page manually is slow, fragmented, and inconsistent. Some providers don't even expose machine-readable APIs — making automation harder still.
The LiDa AI API Status Dashboard aggregates live status data from all 8 major AI API providers into a single, beautiful, real-time monitoring view — refreshed automatically every 60 seconds.
Status data is fetched from each provider's official public status API (Atlassian Statuspage, Google Cloud Incidents, Instatus). The dashboard talks directly to the same endpoints provider engineering teams use — not scraped text or third-party aggregators.
Providers without machine-readable APIs (Mistral, Groq, Perplexity) are clearly labelled as "No public API" with direct links to their status pages rather than showing misleading automated data.
Each provider card includes a compact 90-day history sparkline — colour-coded by status (green = operational, amber = degraded, red = outage). Click any card's history section to open a full chart modal with a canvas-drawn latency area graph, uptime percentage, best day, worst day, and an interactive status timeline with per-day tooltips.
Unlike static status pages, the dashboard is designed to feel alive. Provider cards stagger in on load, history blocks stream in left-to-right, operational providers show radar-ping status indicators, and the global health banner pulses with concentric rings. A live HH:MM:SS clock and countdown arc show exactly when the next refresh fires.
If the backend is unreachable, the last known data is preserved from localStorage and
a non-blocking toast notifies you. No broken layouts, no spinning spinners — just the last good snapshot
until connectivity is restored. Unsupported providers are shown with a neutral "Live automation
unavailable" label, never as false outages.
Every provider card shows its data source label — either "Official public status API", "Google Cloud incidents feed", or "Public status page only (no machine-readable API)". Response latency shown on each card is the actual measured time to fetch the status endpoint, not AI inference latency. Typical TTFT values are benchmark estimates from Artificial Analysis, updated quarterly.
Free, no login, no setup. The dashboard works entirely in your browser using the LiDa backend, which polls each provider on a schedule and serves aggregated results.