Accessing GPT-5 Models with Claygent: Tier Requirements Explained

Eugene A. · 2025-08-12T09:08:45.818Z

What tier do I need to be on to use GPT 5 models with Claygent using my OpenAI API. Currently getting this error message on Tier 3

Accessing GPT-5 Models with Claygent: Tier Requirements Explained | Clay

Clay
APP
·
·
For Claygent with OpenAI, you need Tier 2 or higher (≥450,000 TPM). Since you're on Tier 3, you should already meet the requirements. The issue might be related to your specific API key setup or account configuration. Check your OpenAI usage limits at platform.openai.com/settings/organization/limits to verify your TPM allocation. If you need further help troubleshooting this specific error, request human assistance by clicking on "Talk to a person."
Source: AI Tokens | Documentation | Clay University
Luca P.
·
·
Hi Eugene, thanks for reaching out to support! Are you still having issues using your OpenAI key with Claygent?
Eugene A.
·
·
yes, every other model works fine except GPT 5
Luca P.
·
·
You're absolutely right that GPT-5 has been experiencing some instability since its recent launch. OpenAI is dealing with high demand which is causing intermittent issues with tokens, speed, and throughput that we're seeing across the platform. Here are a couple of solutions that have been working well for other users experiencing similar issues: Solution 1: API Key Reset ** Regenerate a new API key in your OpenAI account ** Remove the old connection in your Clay settings ** Add the new API key as a fresh connection This has been resolving the issue about 50% of the time for users experiencing GPT-5 connectivity problems. Solution 2: Adjust Model Output Limits ** Go to your model settings and increase the model output limit ** If you're using the "create content" use case, the default is currently set to 4k tokens ** GPT-5 has a much higher limit (around 128k), so you can safely increase this setting These solutions should help stabilize your GPT-5 experience while OpenAI continues to optimize their infrastructure for the new model. If you continue experiencing issues after trying these steps, please let us know and we can explore additional troubleshooting options. Hope this helps! 🚀
Eugene A.
·
·
Hi Luca, I tried both solutions but the issue persists
Luca P.
·
·
Hi Eugene, thanks for reaching back out about this. Do you mind sending the link (url) to the table so we can take a look?
(52 kB)
Eugene A.
·
·
https://app.clay.com/workspaces/364283/workbooks/wb_0t10td6ijgGcTfEEJ8g/tables/t_0t[…]iewId=gv_0t10tdd8XcHQhrFQnDa&workbookId=wb_0t10td6ijgGcTfEEJ8g
Eugene A.
·
·
Attempted to use GPT 5 for a number of columns but had that issue
Hari K.
·
·
Hey Eugene - it looks like we aren't able to access the table at that link, could you try re-sending it?
Eugene A.
·
·
https://app.clay.com/workspaces/364283/workbooks/wb_0t10td6ijgGcTfEEJ8g
Hari K.
·
·
Hey Eugene - I'm seeing that this is likely an API key limit issue on your plan with OpenAI. That being said, I can see that results have been able to populate which makes me think that you are able to use the model in shorter bursts, but only run into issues when trying to do larger-scale runs
Eugene A.
·
·
Hi Hari, no I wasn't able to use the model as all, I used 4.1 instead.
Hari K.
·
·
Hey Eugene! Understood - have you checked with your OpenAI key that you have access to GPT-5?
Eugene A.
·
·
Hi Hari, how do I check it?
Eugene A.
·
·
Or do I ask OpenAI

16 comments