Live status · AI
Is OpenAI API down?
Up · checkedNo, OpenAI API is not down right now: api.openai.com answered our check at 23:16 UTC, and OpenAI API’s status page reports no problems.
The OpenAI API gives developers access to GPT models for their own apps.
14-day free trial · Then free on 5 monitors · No credit card
Up
OpenAI API is up
api.openai.com answered our last check in 200 ms. The official status page reports no problems.
- Last checked
- Response
- HTTP 401 · 200 ms
- Checked from
- New York, Stockholm and London
- Uptime since 16 Oct
- 99.985%
HTTP 401 is expected here: api.openai.com asks our check for a key, which shows it is answering.
Official status: All Systems Operational
Status pageEmail me when OpenAI API goes down
One email when it looks down, one when it’s back. Free, no account: confirm by email, stop with one click.
//_01 User reports
Having problems with OpenAI API?
Tell us what isn’t working. Reports are anonymous and show up here within about a minute, so everyone can see whether OpenAI API is failing for others too.
What isn’t working?
Reports, last 24 hours
Per 15 minutes, in UTC
- Last hour
- 0
- Usually per hour
- 0
- Last 24 hours
- 0
No problems with OpenAI API reported in the last 24 hours.
//_02 Common problems
OpenAI API not working? Common problems and fixes
What usually helps when OpenAI API fails for you while it works for others.
Check api.openai.com from our serverGuides
429 errors: rate limit or “insufficient_quota”
A 429 means your project hit its rate limit, or has no credit left (insufficient_quota). Retry with exponential backoff, and check the limits and billing for the project in the OpenAI dashboard.
500, 502 or 503 errors
These come from OpenAI’s side. Retry with backoff; if most requests fail, check the status page above and fall back to another model or provider until it recovers.
401 Unauthorized
The key is missing, wrong, revoked or belongs to another project. Our own check gets a 401 on purpose: it calls the API without a key, which shows the API is answering.
Slow answers or timeouts
Big models and long outputs take longer. Stream responses and give your HTTP client a timeout of a minute or more; slowness across every model at once points to an incident.
//_03 Last 24 hours
OpenAI API response time
How long api.openai.com took to answer our checks from New York, Stockholm and London, averaged per 15 minutes.
- Average
- 159 ms
- Slowest
- 226 ms
- Failed checks
- 0
//_04 Uptime
OpenAI API uptime, last 90 days
- 24 hours
- 100%
- Since 16 Oct
- 99.985%
//_05 Outages
Recent OpenAI API outages
Since we started checking OpenAI API on 16 Oct 2026, our checks found OpenAI API down 2 times, for 4 min in total. The longest outage lasted 2 min, on 23 Oct 2026. That is 99.985% uptime from our checks. OpenAI API’s status page reported one incident in that time, on 21 Oct 2026.
- Outages in our checks, since 16 Oct
- 2
- Downtime, since 16 Oct
- 4 min
- Longest outage
- 2 min
- Status page incidents, since 16 Oct
- 1
Outage timeline
-
Timeout after 10 s
23 Oct 2026, 15:54 UTC · 2 min · our checks
-
Timeout after 10 s
23 Oct 2026, 15:50 UTC · 2 min · our checks
-
Elevated FedRAMP API error rates for Astra and Sol
21 Oct 2026, 14:19 UTC · 28 min · status page
Open on OpenAI API’s status page
-
ChatGPT Work and Dots Errors
Monitoring
OpenAI API components on its status page
- Chat Completions Operational
- Responses Operational
- Agents Operational
- Decisions Operational
- Fine-tuning Operational
- Embeddings Operational
- Images Operational
- Batch Operational
- Audio Operational
- Moderations Operational
- Realtime Operational
- Files Operational
- Login Operational
- Live Operational
- Conversations Operational
- Login Operational
- ChatGPT Work Operational
- Codex in ChatGPT Desktop Operational
- Compliance API Operational
- Search Operational
- File uploads Operational
- Voice mode Operational
- GPTs Operational
- Image Generation Operational
- Deep Research Operational
Every OpenAI API outage since our checks started, with start, end and duration, and an Atom feed for your reader or Slack.
Show OpenAI API’s status on your site
Does your product depend on OpenAI API?
Uptimecheck watches OpenAI API’s status page and your own site, and alerts your team by email, Slack, Microsoft Teams, Discord, Telegram, push or webhook when either has an outage, so you can tell your customers first.
//_06 Is it just you?
Check any website right now
One request from our server in Stockholm: if it answers there, the problem is probably your network.
Checked from Stockholm, Sweden · Free, no sign-up
More on api.openai.com: SSL certificate · DNS records · email setup · domain expiry
About OpenAI API outages
Is OpenAI API down right now?
No, OpenAI API is not down right now: api.openai.com answered our check at 23:16 UTC, and OpenAI API’s status page reports no problems.
When was OpenAI API last down?
Our checks last found OpenAI API down on 23 Oct 2026 at 15:54 UTC, for 2 min: Timeout after 10 s. OpenAI API’s status page last reported an incident on 21 Oct 2026: “Elevated FedRAMP API error rates for Astra and Sol”.
How often does OpenAI API go down?
Since we started checking OpenAI API on 16 Oct 2026, our checks found OpenAI API down 2 times, for 4 min in total. The longest outage lasted 2 min, on 23 Oct 2026. That is 99.985% uptime from our checks. OpenAI API’s status page reported one incident in that time, on 21 Oct 2026.
How does Uptimecheck check OpenAI API?
We request https://api.openai.com/v1/models every 2 minutes from New York, Stockholm and London. Any answer below HTTP 500 within 10 seconds counts as up, since large sites often answer automated requests with 403. A failure only counts once a second check confirms it. We also read OpenAI API’s official status page every 2 minutes.
What can I do when OpenAI API is down?
Check OpenAI API’s official status page for updates and try again in a few minutes. If api.openai.com works for us but not for you, try another network or browser, or clear your DNS cache. If your own product depends on OpenAI API, tell your customers on your status page before they email support.
Why is OpenAI API not working for me when it is up?
Our check shows that api.openai.com answers. Your own network, a VPN, browser extensions, an outdated app or a problem with your account can still break OpenAI API for you alone. The common problems on this page list what usually helps, and when many people report the same problem, it shows here.
Can I get an alert when OpenAI API has an outage?
Yes. Start a free trial and Uptimecheck watches OpenAI API’s official status page for you and alerts you by email, SMS, Slack, Microsoft Teams, Discord, Telegram or webhook when it reports an outage, and again when it is resolved.
Know before your users ask.
Uptimecheck checks your own websites, APIs and cron jobs as often as every minute, confirms every failure from a second location and alerts you before your users start searching.
14-day free trial · Then free on 5 monitors · No credit card