Deploy Streaming TTS on SageMaker with vLLM-Omni

This tutorial shows how to deploy Qwen3-TTS on Amazon SageMaker AI using the AWS vLLM-Omni Deep Learning Container to build real-time voice applications. The setup streams text in and audio out over a persistent bidirectional WebSocket connection, allowing speech playback to begin before the full response is generated. A Gradio application demonstrates the workflow. The post is Part 1 of a series covering specialized DLCs including vLLM-Omni, WhisperX, and llama.cpp. It pairs with a prior post covering the speech-to-text input path using Voxtral-Mini-4B.