Heyy Hanna High latency usually happens because of the speech-to-text model being used, too many languages in one setup, server region distance, or extra steps in the workflow. For Chinese, English, and Spanish, I’d recommend using a faster multilingual model with language detection, or routing each caller to the right language model to keep response time low. What speech-to-text provider are you using right now, and what kind of delay are you noticing?