Skip to content
English
Open Studio
  • There are no suggestions because the search field is empty.

Streaming API

Murf’s Streaming API enables real-time text-to-speech generation for applications that require fast, responsive voice experiences. It is well suited for use cases such as conversational AI, voice agents, virtual assistants, live interactions, and other real-time applications.

 

With Falcon, audio is generated and streamed as the text is processed, enabling low-latency playback and faster response times.

Get started with the Streaming API

You can use the Murf Dashboard to generate Streaming API code based on your selected voice and settings, or integrate the API directly using the Streaming Speech endpoint.

WebSocket Streaming

WebSockets provide the communication channel between your application and Murf for sending TTS requests and receiving audio responses.

Murf TTS API supports WebSocket streaming, which enables low-latency, bidirectional communication over a persistent connection. Instead of sending separate HTTP requests for each piece of text, your application can maintain a single WebSocket connection and continuously send text while receiving synthesized audio.

This approach is particularly useful when text is generated or received incrementally and needs to be converted to speech in real time.

Why use WebSocket streaming?

With a single WebSocket connection, you can:

  • Stream text to Murf as it becomes available.
  • Receive synthesized audio continuously.
  • Reduce the overhead associated with repeated HTTP requests.
  • Deliver faster and more responsive audio playback.
  • Build natural, real-time conversational experiences.

New: Pass model = falcon-2 to use our Falcon 2 model in text-to-speech streaming endpoints, designed for ultra-low latency (~100 ms).

 

With a single WebSocket connection, you can stream text input and receive synthesized audio continuously, without the overhead of repeated HTTP requests. This makes it ideal for use cases where your application sends or receives text in chunks and needs real-time audio to deliver a smooth, conversational experience.