thenotoriousllama
03/24/2025, 5:28 PMVapi Support Bot
03/24/2025, 5:30 PMtransferPlan
object to your transferCall
tool with the appropriate mode:
json
"transferPlan": {
"mode": "warm-transfer-with-message",
"message": "Hey, this call has been forwarded through Vapi."
}
### 2. Request-Start Message Interference
The request-start
message might be interfering with your custom message. Try setting it to an empty string:
json
"messages": [
{
"type": "request-start",
"content": "",
"contents": [],
"conditions": []
}
]
### 3. Message Timing Issues
If the warm message starts too early (before the receiving agent answers), use the warm-transfer-wait-for-operator-to-speak-first-and-then-say-message
setting, which delays the message until after the operator speaks.
### 4. Ensure Text Content is Voiced First
A property named isPrecededByText
can ensure that text content is voiced before the tool call is executed. This might help if the message and transfer are happening too close together.
### 5. Model Configuration
Try setting the model to "gpt-4o/gpt-4o-mini" and adjusting the temperature to 0.3, as this has resolved similar issues for other users.
### 6. Check for Empty Message Field
An empty message field in the transfer tool configuration will result in a silent transfer. Make sure your message field contains text.Vapi Support Bot
03/24/2025, 5:30 PMKings Big
03/24/2025, 11:32 PM