OpenAI favicon

OpenAI outage on 2026-11-05

openai.com
November 2026 5 Minor incident

ChatGPT and Codex ran into elevated errors for 13h 6m on 5 November 2026, under an incident OpenAI labelled "ChatGPT Work and Dots Errors". The fault sat on OpenAI's side, not yours.

Started

15:25 UTC

Duration

Lasted 13h 6m

Source

IsDown

Next time

OpenAI is degraded. Is your product working?

An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.

50 monitors for your own websites. No card.

Affected components: Codex, ChatGPT

What happened?

OpenAI opened the incident at 15:25 UTC, saying it was investigating a potential issue for the listed services and checking how it affects users. At 16:57 UTC it identified elevated errors affecting the impacted services and said it was working on a mitigation. By 17:05 UTC a mitigation was applied and OpenAI moved to monitoring the affected services to confirm recovery. The incident title names Codex and ChatGPT as the two components affected, and both stayed listed until the incident closed. The day's duration is given as 13h 6m, running from 15:25 UTC to resolution, even though the last published update was logged at 17:05 UTC. That gap between the last status text and the closing time is on OpenAI's own page, not an error in this account. OpenAI filed three updates in total for the day, and all three are shown above. No user reports came in during this window, so there is nothing from outside OpenAI's own account to compare against. OpenAI's status history shows 101 incidents in the past 90 days, with a median duration of 1 hour 42 minutes, which puts this one well past typical length. Across all time OpenAI has logged 994 incidents, averaging 15.3 a month. No post-mortem has been published, so the cause of the elevated errors is not on the record.

Learning

OpenAI's incident sat inside its own application layer, the kind of fault that shows up as elevated errors inside ChatGPT and Codex rather than as a dead server or a failed network path. An external check that pings OpenAI's endpoints for a response can come back green while people using the product are still hitting errors on specific requests, because the service is technically up and partially broken at the same time. That is exactly the shape this incident took: identified errors, a mitigation applied, then a wait to confirm recovery, all without the endpoint itself going dark. Watching your own product's calls to OpenAI would catch the symptom on your side but would not tell you whether the fault originates with OpenAI or with your own integration. Over the past 30 days UptimeRobot logged 38 such incidents for OpenAI, all resolved, with a median duration of 2h 2m, so this one running 13h 6m sits above the typical pattern. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.

4.7
stars out of 5
284+ reviews on

Start monitoring in 30 seconds.

There's nothing to install. No credit card required. 50 monitors for free.