### Answer For low latency, consider using the faster/lighter model variants available in Retell AI. The available text LLM options include: [(1) ]() -

gpt-4.1-mini

-

gpt-4.1-nano

-

gpt-5-mini

-

gpt-5-nano

-

gemini-2.5-flash-lite

The "mini" and "nano" variants, as well as "flash-lite" models, are typically optimized for speed over capability. If you want to use speech-to-speech models (which can reduce latency by eliminating the text intermediary step) , the options are: -

gpt-4o-realtime

-

gpt-4o-mini-realtime

-

gpt-realtime

-

gpt-realtime-mini

Additionally, you can enable

model_high_priority

set to

true

to use a high priority pool with more dedicated resources for lower and more consistent latency (note: this typically comes with higher cost) . [(2) ]() Retell AI achieves response times as low as 800 milliseconds with optimized configurations. [(3) ]()

​ To learn more, see the sources I considered: [[1] List Retell LLMs](https://docs.retellai.com/api-references/list-retell-llms) [[2] Update Retell LLM](https://docs.retellai.com/api-references/update-retell-llm) [[3] Why Low Latency Matters for AI Agents | Retell AI](https://www.retellai.com/blog/why-low-latency-matters-how-retell-ai-outpaces-traditional-players) ​

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