# Understanding audio streaming



> Stream audio as it generates to reduce latency.



Rather than waiting for a full audio file, you can stream audio chunks as they are generated. Streaming dramatically reduces time-to-first-byte, which is essential for real-time and conversational use cases.

- Use the streaming endpoint or SDK stream helper.
- Play chunks as they arrive instead of buffering the whole file.
- Choose a low-latency model only when that model is exposed by the configured runtime and has been benchmarked for the use case.



---

Powered by MobDial Agents (MobAgents).