Get told when OpenAI breaks
One email when OpenAI reports an incident, one when it is resolved. No account, nothing else.
Used only for OpenAI status alerts. Every email carries an unsubscribe link.
Open OpenAI incidents
- minor
Elevated latency in the Responses API
Ongoing
OpenAI components
24 of 25 operational
Affected right now
- ResponsesDegraded performance
Show all 25 components
- ImagesOperational
- ResponsesDegraded performance
- Codex in ChatGPT DesktopOperational
- LoginOperational
- AudioOperational
- Compliance APIOperational
- FilesOperational
- FedRAMPOperational
- SearchOperational
- SoraOperational
- Ads APIOperational
- Fine-tuningOperational
- ChatGPT AtlasOperational
- Deep ResearchOperational
- RealtimeOperational
- Voice modeOperational
- BatchOperational
- EmbeddingsOperational
- ModerationsOperational
- Codex WebOperational
- Connectors/AppsOperational
- VS Code extensionOperational
- Codex APIOperational
- SitesOperational
- AgentOperational
Recent OpenAI incidents
The most recent incidents OpenAI has published, newest first.
- minor
Elevated errors in ChatGPT conversations for Free and Go plans
Resolved in 22m
- major
ChatGPT Work seeing elevated errors and latency
Resolved in 5h 24m
- minor
Users may experience an increase in error rates in Workspace Agents and ChatGPT Work on Web and Mobile
Resolved in 4h 14m
- minor
Unexpected logouts for some ChatGPT web users
Resolved in 3h 27m
- minor
Elevated Codex API authentication errors
Resolved in 53m
- minor
Elevated Errors for Thinking mode in ChatGPT
Resolved in 2h 56m
- minor
Chatgpt.com is down - all signups and logins are down as of right now
Resolved in 52m
- minor
Elevated errors deploying Sites
Resolved in 20m
- minor
Elevated errors in ChatGPT conversations for Free users
Resolved in 8h 39m
- none
Error while creating custom RBAC roles for Enterprise users
Resolved in 8h 23m
OpenAI status: common questions
- Is OpenAI down right now?
- OpenAI is reporting minor issues. 1 of 25 components is not fully operational. 1 incident is open. This page mirrors the official OpenAI status feed. PingView last read it on September 1, 2026 at 10:15 AM UTC.
- Where does this OpenAI status data come from?
- Directly from the official OpenAI status page at status.openai.com. PingView polls that feed, records every incident it publishes and keeps a permanent, dated history that the vendor page itself does not retain.
- How often is this page updated?
- PingView reads the official OpenAI feed and rebuilds this page every 5 minutes. The timestamp of the last successful read is shown next to the status, so you can tell how fresh the answer is.
- OpenAI reports everything as operational, but my integration is failing. What now?
- A vendor status page is a global signal, not a per customer one. Partial failures, regional routing problems and issues that hit a single account often never appear on it. Check the OpenAI endpoints your own product calls, from more than one region, and compare your timeline against the vendor's. That is what PingView does, and the SLA calculator turns the resulting downtime into a number you can put in front of a client.