The 750s (12.5 minutes) time limit applies to the entire workflow execution, not individual steps. If the total execution time of your workflow exceeds this limit, it will time out. To troubleshoot the issue, you can: 1. Monitor the execution time of each step in your workflow to identify any bottlenecks or slow steps. 2. Check the logs for any errors or unexpected behavior in the steps leading up to the GPT3.5 16K API calls. 3. Add console.log statements in your custom code steps to gather more information about the state of your workflow during execution. If you still face issues or suspect a problem with Pipedream’s handling of the API calls, please visit https://pipedream.com/support for more support options.