Reported by @plucky panther
Bug Report: Token Limitation and Model Performance Degradation
`Steps to Reproduce`
I can no longer send messages over 16k tokens, while earlier today I was able to send messages with over 100k tokens to GPT-4O-mini. Around 10:50 AM (Brasília time), after processing some long texts for summarization, the token count dropped significantly, and now I’m restricted to sending smaller messages.
I first noticed the issue at 10:50 AM (Brasília time) today.
1 - Try to send messages over 16k tokens to GPT-4O-mini and observe the token limit drop.
`Expected Result`
I should be able to send messages exceeding 100k tokens without restriction.
`Actual Result`
"The message you submitted was too long, please reload the conversation and submit something shorter.
`Environment`
Browser: Google Chrome OS: Windows 10 Plus subscriber