Slow OpenAI results are usually due to API tier limits. If you see a yellow warning, your account has low limits which causes slower performance.
For reliable OpenAI use in Clay, you need 30,000 TPM for content generation columns and 450,000 TPM for Claygent/web research. You can check your current limits at OpenAI API Usage Limits.
Consider upgrading your OpenAI API tier for faster performance, or try processing data in smaller batches to work within your current limits.
What type of OpenAI column are you running - content generation or Claygent?
Tier 5 should definitely handle both content generation and Claygent without issues - it has way more than the required TPM limits.
Since you're on a high tier, the slowness might be due to other factors like batch size or network connectivity. Try processing smaller batches (maybe 50-100 rows at a time) to see if that improves speed.
What type of OpenAI column are you running that's slow - is it a "Generate text" column or Claygent web research? [1]
If you still need help improving the speed of your OpenAI results, I’m here to assist. Would you like to provide more details about the specific column or batch size you’re using so we can explore further options?
The latest updates include GPT-5 rollout in Claygent with 8-17% better research accuracy and 38% fewer tokens used, plus new AI models like Claude Sonnet 3.7 and Gemini 2.0. [1]
We also added flexible AI credit budgets - you can now set spending limits per row in any AI column, and get clear error messages when limits are reached. [2]
If you still need additional help with improving the speed of your OpenAI results, I’m here to assist. Would you like to share more details about the specific column or batch size you’re using so we can continue troubleshooting?
