Hi there! 👋 I can see you're having issues with your audio download pipeline. The URL you shared does indeed look different from the standard Vapi audio URL format. Based on the information I found, the current standard Vapi audio URL format uses

storage.vapi.ai

as the base domain, not

voice.skillsociety.workers.dev

. Here's what a current Vapi audio URL typically looks like:

https://storage.vapi.ai/2d4f5efc-494f-436e-a0c6-9ee23afa77ee-1740140195334-2f7b8057-96ad-4d0f-84c5-a5c99eec68cd-mono.wav

To fix your pipeline, you should access the audio URL through the Artifact interface using

call.artifact.recordingUrl

. Here's a code snippet that demonstrates how to properly download recordings:

ts
const CallDownloader = ({ callId }) => {
  const downloadRecording = async () => {
    // Fetch call data to get artifact URLs
    const call = await vapiClient.calls.get(callId);

    if (call.artifact?.recordingUrl) {
      // Trigger browser download
      window.open(call.artifact.recordingUrl);
    }
  };
}

According to [Discord discussions](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1342498699615539222), you should ensure your assistant's artifactPlan is configured with

recordingEnabled: true

to enable recordings. For troubleshooting your pipeline, I'd recommend: 1. Double-checking your API keys and access permissions 2. Reviewing your code for recent changes 3. Testing in a different environment 4. Using debugging tools to trace the request/response flow Is there anything specific about your implementation that you'd like me to help with? Sources: - [Discord Thread 1342498699615539222](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1342498699615539222) - [Discord Thread 1341449578100031488](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1341449578100031488)

Hey Alberto, can you add a description to your ticket?