Watch our router pick the fastest provider, in real time.

Requesty routes every latency-optimized request with Thompson Sampling: one draw from each provider's live latency distribution, lowest draw wins. Sending everything to one provider is the naive strategy this replaces; the traffic split you see is the router exploiting the fastest option while auditing the rest. Refreshed every minute from production traffic. This applies to requests using our latency routing policy: enable it on your API key to get routed like this.