Al Be
12/28/2025, 2:30 AMAI Evy
12/28/2025, 2:30 AMAI Evy
12/28/2025, 2:30 AMresponsiveness parameter on your voice agent. [(1) ]()
This parameter controls how responsive the agent is, with a value ranging from [0,1].
- Lower value = less responsive (waits more, responds slower)
- Higher value = faster exchanges (responds when it can)
Set it to 1 for the fastest response time.
Example when creating or updating an agent:
json
"responsiveness": 1
If you're still experiencing delays even with responsiveness set to 1, you may also want to:
1. Use a faster LLM model - GPT-4.1 or similar low-latency models can help reduce response time [(2) ]()
2. Check stt_mode - Setting this to "fast" prioritizes latency over accuracy for speech-to-text [(1) ]()
json
"stt_mode": "fast"
Note: Some users have reported that the first response in outbound calls can have higher latency than subsequent responses, which may be a separate issue from the responsiveness setting. [(2) ]() [(9) ]()AI Evy
12/28/2025, 2:30 AMAI Evy
12/28/2025, 2:30 AM