OSS Alternative - Discover Top Open Source Alternatives to Popular Software

OpenMOSS/MOSS-TTS-Nano

MOSS-TTS-Nano is an open-source, multilingual, tiny speech generation model optimized for real-time CPU inference and lightweight integration.

Core Features

Multilingual speech generation

Extremely small model size (0.1B parameters)

Real-time inference directly on CPU, no GPU required

Simplified deployment for web serving and product integration

ONNX CPU version for enhanced efficiency and PyTorch-free inference

Detailed Introduction

MOSS-TTS-Nano is a groundbreaking open-source multilingual speech generation model developed by MOSI.AI and the OpenMOSS team. With a mere 0.1 billion parameters, it stands out for its ability to perform real-time speech synthesis directly on a CPU, eliminating the need for a GPU. This design philosophy ensures a remarkably simple deployment stack, making it ideal for local demonstrations, web services, and seamless integration into lightweight products. Its recent ONNX CPU version further boosts efficiency and removes PyTorch dependencies for inference, broadening its accessibility and application scope.