AWS adds Qwen voice cloning model to SageMaker JumpStart

AWS says developers can now deploy Alibaba Cloud's Qwen3-TTS-12Hz-1.7B-Base model from SageMaker JumpStart to a managed endpoint that generates speech in a speaker's voice. A request supplies new text, a short reference recording and its transcript; the model produces audio without retraining. It supports 10 languages and can use a voice sampled in one language to generate speech in another. The deployment gives developers a way to keep audio within their AWS environment and scale the endpoint with demand. It requires access to a supported GPU instance; AWS's walkthrough uses one NVIDIA L4 GPU with 24 GB of memory.