Independent internet telemetry

Know what is
actually reachable.

A representative, outside-in view of the APIs and infrastructure your products depend on—shown across six sample cities.

AMS

Amsterdam · 184ms

IAD

Ashburn · 592ms

LHR

London · 201ms

NRT

Tokyo · 448ms

SIN

Singapore · 431ms

SJC

San Jose · 118ms

Company watchlist

Service health at a glance

Updated 12 seconds ago · independent checks

Recent signal

What changed

09:42

Anthropic API latency returned to baseline in San Jose.

Recovered

09:17

Two elevated checks observed for OpenAI Realtime in Ashburn.

Watching

08:51

Baseten inference remained healthy across all six regions.

Stable

Coverage

Measured, not self-reported

266 active checks compare regional latency, protocol behavior, and availability. Official vendor status stays visibly separate.

Infrastructure seen today

Cloudflare→AWS→Vercel

Baseten · Inference

Chat completions

SSE inference.baseten.co /v1/chat/completions

Median548 ms

P95884 ms

P991.18 s

24h uptime99.98%

Checks2,6612,660 passed · 1 failed

Latency

Percentile bands

1H3H24H1W

— P50 median┄ P95Shaded P25–P75

Evidence

Latest checks

Dummy samples · all times local

Time	Region	Result	TTFT	HTTP
10:42:18 AM	SJC · San Jose	Passed	118ms	200
10:42:16 AM	AMS · Amsterdam	Passed	184ms	200
10:42:12 AM	IAD · Ashburn	Passed	592ms	200