Suggest Enhancements for Claygent: Queuing and Rate Limiting | Clay

Suggest Enhancements for Claygent: Queuing and Rate Limiting

Hey Clay, love the product. Only downside is there is no queuing system for Claygent/AI requests if you use your own key. Which means i have to sit and WAIT until everything errors out and consistently rerun my column. Basic queuing and automatic rate limiting would be AMAZING!

4 comments

Channeled
APP
·
·
We don't currently have a built-in queuing system or automatic rate limiting for Claygent/AI requests when using your own API key. When using your own key, you're subject to the rate limits set by your OpenAI account. To manage this, you can: 1. Check your OpenAI API usage limits on their platform. 2. Ensure you have an appropriate API access level that meets the required token per minute (TPM) rates. 3. Consider upgrading your OpenAI plan if you're hitting rate limits frequently. Is there a specific issue you're encountering with rate limits or errors when using Claygent?
Jake G.
·
·
Thanks! Im actyally using claude API though
Channeled
APP
·
·
We don't have a built-in queuing system for Claygent/AI requests with the Claude API, but you can implement rate limiting in your Clay workflow. Here's how: 1. In your Claygent column settings, find the "Define rate limit" section. 2. Set a "Request Limit" (e.g., 5 requests) and "Duration" in milliseconds (e.g., 1000 ms for 1 second). 3. This will control how many API requests are sent within the specified time frame, helping prevent rate limit issues. For Claude-specific actions, we recommend a rate limit of 5 requests per 1000 ms, similar to our Kitt AI integration. If you need further assistance, please request human support by clicking "Talk to a person" in this chat.
Sources: Kitt AI x Clay integration | Clay.com HTTP API with JWT authentication | Documentation | Clay University
Channeled
APP
·
·