Hi, I am experiencing issues with Claude 3.7 sonnet and 4.5 Haiku. I have this prompt running to generate an 8 email sequence. It is quite a big prompt with document context as well. At the moment each cell I run takes ages to load or even errors.
Is this because of my prompt are there currently other issues that cause this?
Yes, there are known issues with Claude 3.7 Sonnet specifically. Recent updates doubled the estimated token usage per run, which makes it more conservative but can cause slower performance and rate limiting, especially with large prompts like yours.
For your 8-email sequence with document context, try:
- Use our Metaprompter to optimize your prompt first - it helps you get better results without multiple iterations that burn credits [1]
- Set a "Clay Credit Budget" in your AI column to control spending per row [2]
- Consider breaking your large prompt into smaller, more focused ones
Before we dig deeper, try using Sculptor (top-right corner) to help optimize your workflow - it's free and might solve the performance issues. Let me know if that helps!
Yes, Claude 4.5 Haiku should handle your prompt better than 3.7 Sonnet. Haiku is designed for faster processing and should be more efficient with large prompts like your 8-email sequence with document context. [1]
Since you're still having issues, try setting a Clay Credit Budget in your AI column to control spending per row - this can help prevent timeouts and errors with complex prompts.
If you still need help with your prompt issues or optimizing your workflow, I’m here to assist. Would you like to provide more details about what you’re trying to achieve or any specific problems you’re encountering?