Rule out network first, then assess provider quality
The toolbox is the evidence layer for the leaderboard and provider directory: when an API is unstable, check IP, latency, official status and browser leak signals first, then return to the provider page for long-term detection data.
Available Tools
Each tool serves evidence verification — not feature padding.
IP Deep ScanReady
Exit IP + ASN + datacenter, proxy/Tor signals + geolocation
Claude Relay VerificationReady
Browser-side trace endpoint check to verify exit IP seen by AI sites
WebRTC LeakReady
Client-side STUN-based WebRTC leak detection
Global LatencyReady
globalping public probes for target RTT (not your local network)
AI Service StatusReady
Real-time aggregation of 11 AI/dev service status pages
DNS LeakReady
DoH multi-resolver DNS leak diagnosis + optional browser-side check
AI API Key Leak RadarReady
Paste OpenAI/Anthropic/Google/Groq/HF keys — browser-side SHA-256 match against 10 public leak databases (raw key never uploaded)
Canvas/WebGL/Audio FingerprintReady
Pure client-side Canvas/WebGL/AudioContext/font/screen fingerprint entropy — zero backend, zero IP exposure
AI Service ReachabilityReady
Browser-side reachability check for major model API endpoints — isolates local network vs official service vs relay issues
💰 Cost CalculatorMain site
Pick a model + estimate monthly tokens, compare official vs relay pricing. Covers GPT-5 / Claude / Gemini, synced daily.
📬 Price Alerts & Weekly DigestFree
Subscribe to model price-change alerts (email when threshold crossed) + weekly data digest. Double opt-in, unsubscribe anytime.