Question

ChatGPT: Request too large for gpt-4o in organization org-xxxx on tokens per min (TPM): Limit 30000, Requested 41839

  • 14 June 2024
  • 1 reply
  • 339 views

This post was split from the topic: 

I’m finding that if I use a short prompt, limit it to 1000 tokens output on the openAI gpt-4o model, with a new assistant each time I am receiving the error

 

Failed to create a conversation in ChatGPT

Error from halted execution: Request too large for gpt-4o in organization org-xxxx on tokens per min (TPM): Limit 30000, Requested 41839. 

 

With the requested amount increasing on each test.

 

I have credit, and my request runs fine on the openAI site.

 

Is there some cache or something?

 


1 reply

Userlevel 7
Badge +11

Hi @Chris1000, welcome to the Community! 🎉

Hope you don’t mind but we moved your reply out into a new topic so that we can address it separately as the error you’re running into is different from the one discussed in that other topic.

The ChatGPT rate limits are independent of the API rate limits so that might be why it’s working on OpenAI but not in the Zap (as it performs actions via their API). I’m not 100% sure what tier you’re on with OpenAI but based on the error it looks like you might be on Tier 1 which is limited to 30,000 tokens per minute:
34f81c4e96939974118494b52a3f5d04.png

Given that you’ve set the Zap to be limited to 1000 tokens I suspect that it might be the conversation history that’s being included in the amount. 🤔

In the Zap, can you try setting a different Conversation ID or leave it blank, then retest the action again to see if that gets it working?

Looking forward to hearing from you on this! 

Reply