Abidi
11/18/2025, 10:19 PMVapi Support Bot
11/18/2025, 10:20 PMstartSpeakingPlan.smartEndpointingPlan property from your configuration and set your transcriber to Deepgram Flux. Here’s an example JSON snippet:
json
{
"transcriber": {
"provider": "deepgram",
"model": "flux-general-en",
"language": "en",
"eotThreshold": 0.7,
"eotTimeoutMs": 5000
}
}
This ensures that Flux's end-of-turn detection is used without interference from Vapi's endpointing logic. For more details, see the [changelog](https://docs.vapi.ai/changelog/2025/5/3).
Source:
- [Changelog May 3, 2025](https://docs.vapi.ai/changelog/2025/5/3)Chiranjeet Mishra
11/20/2025, 10:08 PM