I seem to have a recurring issue. I am attempting to retrieve previous caller information before the assistant speaks to move the conversation along faster and have it be more natural. I have an API call as my starting point, which fires just fine and grabs the caller's information. All of that works great, but the initial message after the API call gets repeated. I have tried using multiple models, temp, basically changed a lot of variables one by one to try an isolate the issue, if it were a model problem. No luck Is this normal behavior when using an API call as the starting node? Here's a call ID for reference. e8fc45f4-1d24-451b-b09f-50767643bd19 This seems to happen 100% of the time so far. All the best!

Using an API call as the start node in vapi often causes the initial message to repeat if both the assistant-level greeting and the conversation node’s first message are set so disable one to prevent duplication..

Good to know. I don't think I've tried that yet. Appreciate the help!

Try switching from 11labs to another voice provider.