I’m getting a rate limit error when using gpt-4o (system) in my scenario. I’m on Tier 1 and I’ve used a total of 1489 input tokens and 197 output tokens on 4o today. I’m pretty sure that’s WELL within the limits…
I saw these threads:
I have plenty of credit in my account.
I created a new API key just in case that was the problem.
I can use 4o-mini without a problem (well, except for the input size).
I went to the playground, and I can chat with the system without a problem, so it isn’t a problem with OpenAI.
Also, the message makes me think that Make doesn’t know what the real problem is (‘likely’ is a tipoff to me), and there is no HTTP error code.
Before I open a support ticket, has anyone else figured out what the problem is when this happens and how to solve it?
We could do a better job here with the error message
It means either that you’ve exceeded your available credits (not the case here, as you noted) or that you’ve exceeded your monthly usage limits, or that you’ve exceeded the allowable number of requests or tokens per minute or per day.
OK, I figured it out. I’m getting messges from GMail and the entire thread is being send to the API, including all the image data. So the rate limits are because I exceeded the 30K tokens (it’s 300K bytes, so probably around 45K tokens, though the image data tokenization is probably much more than that).
Maybe that could be added to the error message for that module.
