### Answer Based on Retell AI benchmarks, your metrics could use some improvement: Latency (1050-1550ms) : Production voice AI agents typically aim for 800ms or lower latency to maintain conversational flow. [(1) ]() Pauses longer than 800ms feel unnatural and can break the conversational flow. Retell AI's benchmark shows approximately 780ms average latency, with their target being sub-800ms. Your 1050-1550ms range exceeds the recommended threshold and may cause conversation overlap or reduced trust from callers. [(2) ]() Cost ($0.11/min) : Retell AI's base pricing is $0.07+ per minute for AI Voice Agents. [(3) ]() $0.11/min is reasonable depending on your LLM and TTS choices—ElevenLabs voices cost $0.07/min and OpenAI/Deepgram voices cost $0.08/min on top of base costs. [(1) ]() Token Usage (1844-2084) : This seems high per turn. You may want to optimize your prompts or consider using a more efficient LLM to reduce both latency and cost. Recommendation: For an inbound front desk agent, focus on reducing latency below 800ms—this is critical for natural conversation flow and customer satisfaction. High latency can lead to customer frustration and higher hang-up rates.

​ To learn more, see the sources I considered: [[1] ](https://www.retellai.com/resources/sub-second-latency-voice-assistants-benchmarks) [[2] ](https://www.retellai.com/resources/ai-voice-agent-latency-face-off-2025) [[3] ](https://www.retellai.com/resources/voice-ai-platform-pricing-comparison-2025) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.