If you planned to use ChatGPT today to help you write code, come up with an exercise routine, or conduct some research, you might be out of luck. On Wednesday afternoon, ChatGPT users began complaining about issues with the AI chatbot. Based on reports from Downdetector and OpenAI’s status page, this isn’t an isolated issue.
OpenAI began investigating issues with ChatGPT at around 11:00 a.m. PT on Wednesday. That investigation continued into the afternoon, and as of 1:10 p.m. PT, the company noted that some users were seeing unexpected error messages.
When I visited ChatGPT moments ago to see if I could replicate the issue, I received this error message when I tried to input a prompt: “You’ve reached the current usage cap for GPT-4. You can continue with the default model now, or try again later.”
I don’t pay for ChatGPT Plus, so I don’t have access to GPT-4. I am attempting to use GPT-3.5, but there’s clearly something busted behind the scenes. There was also another error message: “We’re experiencing exceptionally high demand. Please try again later.”
Earlier today, OpenAI issued a fix for an issue “impacting vision models on ChatGPT” and in the company’s API. It is possible that the fix had an unintended impact on the chatbot, because the elevated errors in ChatGPT started right after the fix was deployed.
At 1:39 p.m. PT, OpenAI shared the following update: “We have identified the source of the elevated errors in ChatGPT and are working through a fix at this time.”