Is OpenAI API down?

Check if openai is down for everyone or just you

CHECKING…

Checking OpenAI API now

- response checked queued

We haven't probed this target yet in the current cycle. The first probe usually completes within 10-30 seconds. Click "Re-check" above to trigger one immediately.

HTTP/1.1strict TLS Pending
-
queued
HTTP/2relaxed TLS Pending
-
queued
Browser UAGET Pending
-
queued
curlHEAD Pending
-
queued
No probe data
Collecting datalast 30 days
See full history →
2026-07-01: no data2026-07-02: no data2026-07-03: no data2026-07-04: no data2026-07-05: no data2026-07-06: no data2026-07-07: no data2026-07-08: no data2026-07-09: no data2026-07-10: no data2026-07-11: no data2026-07-12: no data2026-07-13: no data2026-07-14: no data2026-07-15: no data2026-07-16: no data2026-07-17: no data2026-07-18: no data2026-07-19: no data2026-07-20: no data2026-07-21: no data2026-07-22: no data2026-07-23: no data2026-07-24: no data2026-07-25: no data2026-07-26: no data2026-07-27: no data2026-07-28: no data2026-07-29: no data2026-07-30: no data
30 days agocollecting data · 0/30 daystoday
Response time & status, last 24h
Awaiting first probe
Monitoring data will appear after first check cycle

Notify me

We'll ping you the next time OpenAI API's status flips. One-shot subscription, no account required.

Webhook receives a JSON payload on every status transition. Fire-and-forget — no auth from us; the URL is its own secret.

Monitor OpenAI API — free Get OpenAI API outage alerts — free

One click — the form comes prefilled. A free account saves 5 monitors with email alerts, checked every 5 minutes, plus 3 free vendor watches — the outage-alert list for services you use but don't run. Pro checks every minute — you'll usually know before your users do — and adds webhook + Discord alerts, a public status page, and no ads.

Embed a live status badge for OpenAI API
OpenAI API status badge

Live SVG, refreshed every minute, no signup needed. Shows up / degraded / down straight from our probes.

Methodology

Probe
We send a GET request to the target URL with each probe method and time the full response.
Counts as DOWN
Every probe method fails the request (connect refused, DNS error, timeout, or HTTP 5xx).
Counts as DEGRADED
At least one probe method succeeds but at least one fails, the site is up but blocking or mishandling a specific request type.
Detail
Per-method results include HTTP status code, response time, and best-effort DNS / connect / TLS / first-byte phase timings when the Node HTTP client exposes them.
Cadence
Every 5 minutes, in parallel across 4 independent probe methods (HTTP/1.1 with strict TLS, HTTP/2, a browser User-Agent GET, and a curl-style HEAD) from our edge network. We check probe diversity, not geography — if a site blocks a non-browser User-Agent or only speaks HTTP/1.1, the four cards show it.
Rate-limited targets
If a host returns 429 or consistently drops connections from our IPs, we cap retries at 3 and report the last observed status, we do not flood the target to confirm the outage.
Data source
Our verdict is based on direct multi-method probes, every result above is a live network request, never a Twitter mention or DownDetector aggregate. User reports, when shown, appear separately below as a supplementary signal and are never used to determine the up/down verdict.

isitdown.io is an independent monitoring service and is not affiliated with, sponsored by, or endorsed by OpenAI API. Status data comes from our own direct probes and, where shown, the provider's public status feed.

Common questions

Is OpenAI API down right now?

We have not completed a probe of openai yet. We have not built up enough probe history for openai yet to quote an uptime figure — the numbers above are from the most recent check only. Every figure on this page comes from our own HTTP requests to openai — four independent probe methods (HTTP/1.1, HTTP/2, a browser User-Agent GET and a curl-style HEAD) — not from user reports or a third-party feed.

Is OpenAI API down for everyone or just me?

If all four probe methods above show a response, OpenAI API is reachable from our network and the problem is local to you: flush your DNS cache, switch networks, disable a VPN, or try another browser. If two or more methods failed, the problem is on OpenAI API's side. When every method fails at the connection level we say "unreachable" rather than "down for everyone" — some large sites block monitoring networks at the edge while serving browsers normally, and we would rather be precise than dramatic.

How reliable has OpenAI API been recently?

We do not have enough history for openai to answer that honestly yet. This page fills in as our monitoring accumulates checks; the 30-day strip above shows exactly which days we have data for.