Hi all,
I have used Google AI Studio for a while and not had any issues. Recently though I can not do any long-form work with it because my requests are always denied if I go over an 8000 token limit. I’ve tried with the flash, pro, and flash-8b models, all the same. If I make a request that has more than 8000 tokens in the full conversation, I see the error: “An internal error has occurred.” After looking at the devtools, I find that the error is, for example:
[
,
[
3,
"The input token count (9561) exceeds the maximum number of tokens allowed (8000)."
]
]
In the side-bar it shows the 1,000,000 token limit, in this case actually it displays “11,433 / 1,000,000”.
Have I not provisioned something in my account correctly? I’m on the free-to-use version, but that doesn’t say anything about token limits per requests, only net tokens per minute. And I have used this for months going way above 8000 tokens and it never gave errors before.
Any thoughts?
Thanks!