Baidu Text-to-Speech vs GPT-SoVITS

Baidu Text-to-Speech converts text into speech for applications and devices, with Chinese, English, mixed Chinese-English, and more than 70 voice styles available through its online service. Users can change speech rate, pitch, volume, and supported emotional effects; Chinese polyphonic characters can be annotated with pronunciation. Streaming synthesis can play audio nearly as text is entered, and long-text synthesis handles up to 100,000 Chinese characters in one request. Baidu provides HTTP and WebSocket APIs, online and offline SDKs, and Android and iOS SDKs for phones and smart devices. Offline synthesis is intended for settings with weak or no network access, including mobile apps, readers, robots, and smart hardware. Listed uses include audiobook reading, news and order announcements, and smart hardware. Short-text API requests must stay under 1,024 GBK bytes, with 120 bytes or less recommended. Export formats are MP3, PCM, and WAV. The service offers a free trial; offline trial and license plans are listed, while prices for other listed plans are not provided.

GPT-SoVITS is a free WebUI for few-shot voice conversion and text-to-speech. It can generate speech from a five-second vocal sample, or use one minute of training data to improve voice similarity and realism. Cross-lingual inference supports English, Japanese, Korean, Cantonese, and Chinese. The WebUI includes tools for separating accompaniment, segmenting training sets, multilingual speech recognition, and labeling text. Speech-recognition options include Fun-ASR-Nano, SenseVoice, classic FunASR, and Faster Whisper. The repository also provides an API with GET and POST inference endpoints that return WAV audio streams on success. Installation instructions cover Windows, Linux, and macOS, with Docker Compose services in full and Lite variants. The Lite image omits ASR and UVR5 models; UVR5 models must be downloaded manually, while ASR models download as needed. On macOS, the project temporarily uses CPUs because it reports significantly lower quality from models trained with Mac GPUs. GPT-SoVITS is released under the MIT License and links to Chinese and English user guides.