🔵 开发中 Dev
Overview
Real-time monitoring of global major LLM providers (OpenAI/Anthropic/Google/Groq), tracking uptime, Time-To-First-Token (TTFT), and historical reliability.
Background
Providing transparent multi-region health and performance benchmarks to help engineers build resilient AI applications.
Key Highlights
- Global multi-node distributed active probing measuring exact DNS, TCP, and TTFT metrics
- Side-by-side comparison across model versions, token throughput, and pricing
- Real-time incident detection and health degradation alerts
Architecture
Distributed edge probing workers aggregating metrics into time-series storage and distributed via edge CDN.
Key Metrics
40+ models
Models monitored — Mainstream commercial & open-source
60 seconds
Probe frequency — High-precision heartbeat polling
365 days
Historical retention — 365 days of SLA verification
Tech Stack
Tech Stack:Next.js, React, TypeScript, Edge Workers, Tailwind CSS
Project Info
Year:2026 · Status:Live