Troubleshooting GPT-5 Connection Error with Clay Support

George D. · 2025-10-19T18:08:22.297Z

Hi @Clay Support I've added a Open Ai connection with Tier 1, we are using GPT5 with 500k TPM , truing to write a prompt but keeps giving me this error. Can you help?

Troubleshooting GPT-5 Connection Error with Clay Support | Clay

Clay Support
APP
·
·
The error indicates your prompt hit the token output limit. You have two solutions:
Option 1: Increase Clay Credit Budget
Edit your AI column and increase the "Clay Credit Budget" to allow more tokens per row. This lets you use additional credits for longer outputs without changing every row's cost. [1]
Option 2: Upgrade OpenAI Tier
Your Tier 1 account may be limiting performance. For reliable AI operations, you need Tier 2 or higher (≥450,000 TPM for advanced features). [2] Check your limits at OpenAI's platform and upgrade if needed.
Try increasing the credit budget first - it's the quickest fix for token limit errors.
Clay Support
APP
·
·
Was that helpful?
George D.
·
·
But I'm using GPT5 which has 500k TPM
Clay Support
APP
·
·
The error isn't about your OpenAI tier limits - it's about Clay's token output limit for that specific prompt. Even with 500k TPM, individual prompts can hit Clay's max_tokens setting.
Increase your "Clay Credit Budget" in the AI column settings. This allows Clay to use more credits for longer outputs when needed, without affecting every row's cost. The budget only applies when prompts require extra tokens. [1]
Clay Support
APP
·
·
Is that what you were looking for?
George D.
·
·
how much is the max tokens value?
Clay Support
APP
·
·
The max tokens value depends on the AI model you're using:
Claude models: Maximum of 4,096 tokens (defaults to 256) [1]
Claygent: Uses a "Max Cost" setting instead of token limits - defaults to $0.50 per request [2]
For your specific error, increase the "Clay Credit Budget" in your AI column settings to allow more tokens when needed.
Clay Support
APP
·
·
Did that answer your question?
George D.
·
·
I'm using my own OpenAi account with GPT5 Mini model
Clay Support
APP
·
·
When using your own OpenAI API key with GPT-4o Mini, the max tokens limit is controlled by Clay's "Clay Credit Budget" setting, not OpenAI's model limits.
Edit your AI column and increase the "Clay Credit Budget" to allow more tokens per row. This lets Clay use additional credits for longer outputs when needed, without affecting every row's cost.
The budget only applies when prompts require extra tokens - it doesn't increase costs for all rows. [1]
Clay Support
APP
·
·
Was that helpful?
George D.
·
·
This error refers to the length of the prompt or of the response?
Clay Support
APP
·
·
The error "max_tokens or model output limit was reached" refers to the response length, not your prompt length.
This happens when the AI's generated response would exceed Clay's token output limit for that specific request. Your prompt can be any size, but Clay controls how long the AI's response can be through the "Clay Credit Budget" setting. [1]
Increase your Clay Credit Budget in the AI column settings to allow longer responses when needed.
Clay Support
APP
·
·
Was that helpful?
George D.
·
·
and what's the max_token for Clay starter?

21 comments