OpenAI enrichment is incredibly slow for me, any1 else seeing this?

Slow OpenAI enrichment is typically due to low API tier limits on your OpenAI account. For reliable performance, you need 30,000 TPM for ChatGPT columns and 450,000 TPM for Claygent - free plans and tier 1 won't support these requirements. Check your API usage tier at OpenAI API Usage Limits and consider upgrading to a higher tier for faster performance. You can also break large requests into smaller batches to help with processing delays.

Hello Noah! Thanks for reaching out! Could you share the table you are facing this in and how recurring this is? We had a bug yesterday around this but it has hence subsided and I have not seen any recent reports on this. We can always take a look!

Hey Noah! I created a duplicate view named 'Support View' to see which cells are queued. I can see about 5.8k rows are queued. How long have these been queued for again? I plan on escalating this to the eng team if it has been stuck in queued for a while so would like all context possible before I do!

Ahh got it. I am going to wait for a few more minutes to observe the speed and escalate this soon. I will keep you updated!

Hey Noah, since i last checked about a 100 rows have been enriched in the past twenty minutes. I will reach out to the eng team to confirm if that is abnormal.

Hey Noah! Thanks for being patient while I got back to you! I just got back from a discussion with the eng team and wanted to mention a few things we noticed and what might help. The main point being we noticed you are currently using your API key for these actions and not Clay's key. The speed at which these enrichments are completed will then depend on the rate limits of your key and slows down when it's close to that limit. I noticed you are not using a very token intensive so that helps. However there are two ways we can go about this: 1. Check your rate limit. If it is close to the rate limit, the speed will reduce. You can contact their support for more information or guidance on how to go about this. You can also think of using multiple API keys, different for different AI columns to spread it across different rate limits and speed it up! 2. The second method would be to use Clay's API key. This will mean credits will be spent on the action but it will definitely speed things up. Since you are currently using your own key, this is not a Clay issue. Hope this helps and do let me know the approach you decide to take!

Does that solve the issue? Is there anything else I can assist you with?

Hey Noah, I am seeing a bunch of reports on this and an ongoing bug, I will go ahead and link this ticket to it. The engineering team is taking a look to see if it is an issue on our end.

Noah - Sorry for this, it looks like we've got a bug on our hands. I've escalated this to our engineering team to help look into it. We'll follow up here as soon as we know more!

Good news! Happy to report that we heard back from the team and the latency issue should be fixed. Let us know if the problem has been resolved on your end, or if you still have trouble. Thank you for your patience as we worked on it!

Hey Noah, thanks for reaching out. So sorry to hear that you are still running into issues. Are you using your own OpenAI API key? I can see that you are being highly rate limited, so you may want to check your usage on the OpenAI API dashboard. Keep me posted on what you see on the dashboard and let me know if you have any further questions!

Hi there - thank you for your patience while our Engineers identified and fixed this issue. We're sending out a confirmation to all customers that reported an issue that everything should be resolved. Please let us know if anything is still not functioning as expected in your account. And thank you again for reporting the issue!