●Incident

OpenAI degradation.

ResolvedDEGRADATION · MEDIUM

Oct 21, 2026, 02:30 AM · about 16 hours ago · Duration: 19min

openai View source

Incident Summary

External incident reported: Degraded performance in ChatGPT, Codex, and APIs.

Timeline

●Started

Oct 21, 2026, 02:30 AM

about 16 hours ago

●Resolved

Oct 21, 2026, 02:50 AM

about 15 hours ago

●Duration

19min

Total downtime

Probecast for agencies

How many of your clients depend on OpenAI?

When OpenAI slows down, every client that depends on it feels it at once, and the first call usually goes to the agency. Probecast watches your clients' sites and the providers underneath them, shows which clients an outage touches, and turns the month into a report your client can trust.

One provider breaks. See every client it touches.

Ordered by how many clients it hits, so you message the right people first.

Each client's site and store, checked from four regions.

US West & East, Europe and Asia — a regional outage never hides.

A monthly report under your brand, reviewed by your team before it goes out.

Uptime, incidents, traffic, revenue and support — signed off by a person.

Your team scoped per client. Every change on the record.

Freelancers see only their clients; every change lands in an append-only audit log.

Book 15 minutes →View OpenAI full status →

Alerts land on email, Slack, push or webhooks — with the client's name on them.