AI DownCLI · API

Is Codex down right now?

No problems detected
1628 msmedian API latency, 30 min
0.0%probe error rate, 30 min
0user reports, 15 min
0terminal users with errors, 15 min
Operationalofficial status

Updated 1 min ago.

Last 24 hours

14:00 UTC: 0 reports15:00 UTC: 0 reports16:00 UTC: 0 reports17:00 UTC: 0 reports18:00 UTC: 0 reports19:00 UTC: 0 reports20:00 UTC: 0 reports21:00 UTC: 0 reports22:00 UTC: 0 reports23:00 UTC: 0 reports00:00 UTC: 0 reports01:00 UTC: 0 reports02:00 UTC: 0 reports03:00 UTC: 0 reports04:00 UTC: 0 reports05:00 UTC: 0 reports06:00 UTC: 0 reports07:00 UTC: 0 reports08:00 UTC: 0 reports09:00 UTC: 0 reports10:00 UTC: 0 reports11:00 UTC: 0 reports12:00 UTC: 0 reports13:00 UTC: 0 reports14:0020:0002:0008:0013:00bars: user reports per hour (max 5) · red line: probe error rate (0–100%)

Report a problem

Is Codex not working for you?

From your terminal

Get Codex errors reported automatically

One command adds a tiny hook to Claude Code and Codex. When a request fails, it sends the error type, tool, model and version. No prompts, no code, no email.

curl -fsSL https://aidown.dev/install | sh

What it does and what it sends · undo with aidown uninstall

Questions

Is Codex down right now?

As of 2026-10-20 13:10 UTC, Codex: no problems detected. Median probe latency 1628 ms, probe error rate 0.0%, 0 user reports in the last 15 minutes, official status: Operational.

How does AI Down check Codex?

We send a small request to the OpenAI API every few minutes from Cloudflare and record errors and latency. We also count reports from users on this page and read the official status page.

Why does the official status page say operational?

Official status pages are updated by hand and often lag behind real problems. Our probes and user reports usually show a problem first.

Other services

Claude · Claude Code · ChatGPT · Gemini · Cursor · GitHub Copilot · OpenRouter