Rate limit reached for gpt-4o-mini in project

support
,
1

1. The problem

I have built my own scripts that use the Open AI API and they work fine, and I use it a lot daily for multiple sites without getting this error.

I went as far as creating a new API key only for SEO CONTENT MACHINE and limiting the usage in the setting to the lowest numbers.

But within 10 seconds of running creating article on only one task, this pops up with these errors about limits being reached.

Is the system maybe sending requests over the API too fast? The usage limit is far away from being reached.

2. Screenshot or task log of the problem

image
image2044×385 89.7 KB

Warning: "{"status":429,"data":{"error":{"message":"Rate limit reached for gpt-4o-mini in project proj_ITNenOqY4PHthXd7gzB5zSrY organization org-dtxGU0tEHFuN8LoMxXJQ2VDj on requests per min (RPM): Limit 3, Used 3, Requested 1. Please try again in 20s.

more.","type":"requests","param":null,"code":"rate_limit_exceeded"}}}" retry 1/3 in 8s

3

I need you to double check your AI threading settings in SCM.

image
image1002×608 27.2 KB

AI article threads = how many articles to create simultaneously
AI prompt threads = how many prompts in that article to run simultaneously

If you change any AI threads, you need to restart SCM app for them to take effect.

4

I even had the task threads at 1 and it still spits out those errors.
Here’s my screenshot changed back to your settings: - I’ll let you know how it goes.

image
image1108×463 16.9 KB

5

Yeah, okay - using your settings that were more aggressive than mine.
I changed the setting to yours - I installed the latest update and started the article creator.

Within less than 1 minute, the errors start.
Here is the log

image
image2755×727 293 KB

6

OK so this is just rate limit from openai, not related too many threads now.

Your account rate limit seems unusually low, 3 requests per minute.

For ai outline with 30 sections it will take 10mins to write that article.

Is there a reason why you have such a low request per minute?

–

Whats also strange is that open ai keeps reporting a wait of 20secs, no matter how much time elapsed between requests by SCM. The requested amount also never drops down from 3/3.

Are you using this openAI account with other tools?

Eg: groqai also have RPM, but its based on tokens and the limit correctly reports your usage over time while on your openai account it looks permanently over used and there is no free slots for SCM to send requests.