Integration Methods

When integrating SESTEK TTS into your application, choosing the right method depends on your architecture, latency requirements, and deployment environment. This page compares the three available integration methods - HTTP, MRCP, and WebSocket - and helps you decide which fits your use case best.

Quick Comparison

HTTP (REST)

HTTP is the simplest and most widely compatible integration method. The client sends an HTTP request containing the text to be synthesized, and the TTS system returns an audio file in response.

Best for: Applications where the text is known in advance, audio can be pre-generated, and real-time streaming is not required.

MRCP

MRCP (Media Resource Control Protocol) is an industry-standard protocol for integrating speech synthesis with voice application servers. It provides standardized, real-time control over the synthesis process and is the recommended integration method for IVR platforms.

Best for: IVR systems and call center platforms where telephony integration and real-time performance are critical.

WebSocket

WebSocket provides a persistent, bidirectional communication channel for real-time, full-duplex audio streaming. Unlike HTTP, the connection stays open - allowing text to be sent continuously and audio to be returned as it is synthesized, with minimal delay.

Best for: Applications that require low-latency, continuous speech output where audio needs to start playing before the full text has been processed.

Choosing the Right Method