Probes from EU & SEA seeing 5xx on a subset of requests. We are tracking the provider's investigation.
Resolved — API timeouts resolved
Brief timeout window on completions endpoint, fully recovered.
Major outage — API & ChatGPT unavailable major
Global 5xx across endpoints. Root cause: upstream provider incident.
✨ AI summary
OpenAI is currently partially degraded: the Chat Completions API shows elevated errors in EU/SEA since 13:40 UTC, while the ChatGPT web app and image endpoints remain operational. Over the past 90 days uptime is 99.86% with 2 notable incidents. If your app only uses the web product, you are likely unaffected.
FAQ
Is OpenAI down for everyone?
Why does dudown say degraded but the app works for me?
Degradation is often regional or endpoint-specific. Our probes saw errors from EU/SEA on the Completions API; if you're on the web app or another region you may be fine.
How often do you check?
Every 30 seconds from 7 regions. The "last checked" timestamp at the top reflects the most recent successful probe.
Are user reports reliable?
They're a helpful early signal but unverified. Our verdict is driven by active probes + the official status page, not raw report counts.
⚠️ Mock / for reference only. dudown is an independent monitor and is not affiliated with OpenAI. Status shown here is fictional sample data. Always confirm critical decisions against the provider's official status page.