Tell HN: Codex Is Down

9 points by minimaxir ↗ HN
Shows an Incorrect API Key error...which is weird. Nothing on the status page.

Per Twitter/X this appears to be widespread

37 comments

[ 0.48 ms ] story [ 8.7 ms ] thread
Friday afternoon in the US I guess, perfect time for things to go wrong.
Looks like someone at OpenAI did a Friday afternoon deployment lol
Excellent work all around, take the rest of the week off.
thank god I'm not the only one, pulling my hair out hitting ctrl+z ... Happy Friday I guess
Codex took an unannounced early day off for the weekend. It happens.
I thought I got banned.
A good time for reading
Or switching to Claude
I switched to Claude. Opus 5.5 was causing fomo anyway. This was the trigger. I admit I have zero loyalty
Somebody really wanted to cross one off the weekly TODO list :)
What happens when your entire day is depended on saas models? Anyone even considers using their locally hosted models in order to finish their work? Do you trust your local models to do any of you outstanding work? Do we even understand the code anymore to manually fix anything?
I was worried somewhat until it become really apparent that small models that run locally can handle 99% of technical computer crap when given a decent tool set harness and net access.

so, to answer your question more practically, during this outage a luna agent of mine fell back to a local bonsai 2 model hosted on my desktop, and the work still got done just fine, just took longer.

unrelated but also chatgpt is super buggy, try generating an image, click stop, now wong be able to send another message until you reload, this issue has been there for months
Nevermind. Switching to Opus 5.5.
I have $200/mo Claude and $200/mo Codex subscriptions and run both agents all day long, so this is just a mild inconvenience.
Way to go, Ace!
They getting ready to release a new model?
It's crazy how fast these things became load-bearing
I don't know if related, but a half hour before it fully went down, it was flagging all of my requests as security violations which was bizarre.
Appears to be back online.
It doesn't help that the error message was entirely incorrect, not an accurate reflection of the problem. In reality the user wasn't using any API key whatsoever.