What is the google API's uptime?

Measured uptime for the google API averages 100% across 4 region(s) over the last 24 hours.

google — AI API latency & reliability by region

Last verified September 07, 2026

As of September 07, 2026, the google AI API responds in 40 ms p50 (edge, time-to-first-byte) from US (Central) — its fastest measured region — with 100% uptime. Slowest is South America (São Paulo) at 160 ms p50.

Is google down right now?

Reachable from all 4 regions, requests to gemini-flash-lite-latest succeeding

Last successful probe Sep 7, 00:17 UTC · every 5 minutes from 4 region(s)

There was an outage: on Sep 4, 19:30 UTC, requests to gemini-flash-lite-latest failed from 2 of 4 regions simultaneously for under one 5-minute probe cycle. No other provider failed in that window, so a shared route or CDN does not explain it. Measured from outside google’s own network, independently of its status page; full log: measured AI API incidents.

Real requests to gemini-flash-lite-latest: 273 of 288 succeeded in the last 24 hours from 4 region(s); last failure Sep 6, 23:05 UTC from South America (São Paulo) (timeout). A one-token completion is sent on the same schedule as the host probe; 5xx, 529 and timeouts count as failures, problems with our own key or quota are excluded.

Google publishes no machine-readable status page, so there is nothing to cross-check against.

Edge latency, 24 h

40 ms

p50 from US (Central), the fastest region

Uptime, 24 h

100%

1,152 probes across 4 region(s)

Failed probes, 72 h

of 3,455 probes, all regions

Failed requests, 24 h

of 288 to gemini-flash-lite-latest

Reported by Google, 72 h

—

no machine-readable status page

Was google down in the last 3 days?

Yes. 3 outage(s) could be attributed to Google in the last 72 hours: probes failed from two or more regions at once while no other provider was failing.

Sep 6, 23:05 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 20:45 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 17:45 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 16:05 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 12:45 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 11:25 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 10:05 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 09:45 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 08:45 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 07:25 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 07:05 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout
Sep 6, 06:05 UTCrequests to gemini-flash-lite-latest: South America (São Paulo) only — not an outage · under one 5-minute probe cycle · 1 failed probe(s) · timeout

What Google’s own status page reported

Google publishes no machine-readable status page, so there is nothing to set against the measurements above.

Is google getting faster or slower?

Right now google is slower than usual from US (Central): 120 ms against a three-day typical of 46 ms.

hourly p50 latencyall probes succeededsome probes failedhalf or more failedno data

One panel per region: the line is the hourly p50 edge latency (shared scale, ms), the strips beneath it are the outcome of that hour’s host probes and, below them, of the real requests to gemini-flash-lite-latest. Hover any hour for the numbers. Flat is good; a rising line means the endpoint is slowing down from that origin, a gap means no probe succeeded in that hour.

How fast is the google API from each region?

Region	p50 TTFB	p95	Uptime
Asia (Tokyo)	63 ms	105 ms	100%
Europe (Germany)	99 ms	202 ms	100%
South America (São Paulo)	160 ms	297 ms	100%
US (Central)	40 ms	107 ms	100%

Measured from US (Central), google responds in 40 ms; from South America (São Paulo) the same API takes 160 ms — 4.0× longer for an identical request. That difference is the network path, not the model, and it applies to every call you make.

What is actually being measured here?

A probe in each region opens a real connection to google’s own API host every five minutes and times DNS resolution, the TCP handshake, the TLS handshake and the first byte of the response. Nothing is routed through a gateway or aggregator, so these figures describe google’s infrastructure rather than a reseller’s. Uptime counts a probe as failed only when the service genuinely fails — a 401 from an unauthenticated probe means the endpoint answered correctly and is counted as up.

What this page does not tell you

These are edge latency numbers: the time before the model begins generating. They say nothing about answer quality, throughput, price, or how long a full completion takes — those depend on the model you call. Edge latency matters because it is unavoidable: it is paid on every request, before a single token exists, and no prompt engineering removes it. Where an API key is available, inference time-to-first-token is measured separately and shown on the region pages with the model named next to the figure. Likewise, an unauthenticated probe cannot see errors that hit only some models or some accounts — the kind a status page usually reports — which is why Google’s own reports are shown next to the measurements rather than folded into them.