1. /tts?t=[text]&v=[voice]&r=[rate]&p=[pitch]&o=[outputFormat] try
2. /voices?l=[locate, like zh|zh-CN]&f=[format, 0/1/empty 0(TTS-Server)|1(MultiTTS)] try
3. OpenAI兼容API: POST /v1/audio/speech (JSON格式: {"model":"tts-1","input":"要转换的文本","voice":"alloy"})