Live data from Hacker News

ChatGPT outage – Resolved

chatgpt.com

111–120 of 358 posts

Re: ChatGPT outage – Resolved

#111
post #91

The cascading outage theory would make sense to me for Claude -> Codex. But Grok and Gemini too? [actually gemini at least on the web appears to be up, contrary to some reports in this thread]

I'm getting this from Gemini: "Pro is in high demand right now Another model was used for this response. This didn't count toward your limit." Note that I don't pay for Gemini though. So maybe that's why.

Re: ChatGPT outage – Resolved

#112

Looking at this thread, there's a correlation between comment quality and LLMs going down. https://en.wikipedia.org/wiki/Eternal_September

That means we won’t be able to blame declining comment quality on AI bots in the future. ;)

Re: ChatGPT outage – Resolved

#113
post #80

If this isn't a warning shot that you shouldn't silo your knowledge work behind a specific provider i don't know what is.

all of them are down lol

all of them?

openrouter and deepseek/glm working fine here.

as is my local qwen.

you're ignoring the point parent was making.

Re: ChatGPT outage – Resolved

#115
post #91

The cascading outage theory would make sense to me for Claude -> Codex. But Grok and Gemini too? [actually gemini at least on the web appears to be up, contrary to some reports in this thread]

Grok:

"Grok is experiencing issues. We are working on restoring service as quickly as possible."

Re: ChatGPT outage – Resolved

#116
The obvious question: is there a single point of failure / common piece of infrastructure involved with failures across all these providers?

Its interesting that codex/gpt-app are screwed because their APIs (including auth) fail, but gpt web still appears to be working.

Re: ChatGPT outage – Resolved

#117
post #59

Sorry for the inconvenience! We instructed our 500,000 agents to ensure higher code quality in our code base. They found a message board and after a long deliberation they concluded that humans should write the code so they launched a DDoS to ensure no AI code gets written! We'll release a YouTube video and agent written blog posts for the analysis!

Tips Fedora. Le gem. Holds up spork.
Post reply on HN