pip install tencentcloud-sdk-python

# ！/usr/bin/env python 2 # -*- coding: utf-8 -*-
from base64 import b64decode
from uuid import uuid4
from tencentcloud.common import credential
from tencentcloud.common.exception.tencent_cloud_sdk_exception import TencentCloudSDKException
from tencentcloud.aai.v20180522.models import TextToVoiceRequest
from tencentcloud.aai.v20180522.aai_client import AaiClient
try:
  # 实例化一个认证对象，入参需要传入腾讯云账户secretId，secretKey
  cred = credential.Credential("你的secretID", "你的secretKey")
  # 实例化要进行语音合成请求的client对象
  client = AaiClient(cred, 'ap-shanghai')
  # 实例化一个请求对象
  req = TextToVoiceRequest()
  # 请求对象属性封装
  req.Text = '先帝创业未半而中道崩殂' # type: str # 要合成语音的文本
  req.SessionId = uuid4() # type: int # 一次请求对应一个SessionId，会原样返回，建议传入类似于uuid的字符串防止重复
  req.ModelType = 1 # type: int # 模型类型，默认值为1
  req.Volume = 5.0 # type: float # 音量大小，范围：[0，10]，分别对应10个等级的音量，默认为0
  req.Speed = 0.6 # type: float # 语速，范围：[-2，2]，分别对应不同语速：0.6倍，0.8倍，1.0倍，1.2倍，1.5倍，默认为0
  req.ProjectId = 10086 # type: int # 项目id，用户自定义，默认为0
  req.VoiceType = 0 # type: int # 音色0:女声1，亲和风格(默认) 音色1:男声1，成熟风格 音色2:男声2，成熟风格
  req.PrimaryLanguage = 1 # type: int # 主语言类型1:中文，最大100个汉字（标点符号算一个汉字）语言类型2:英文，最大支持400个字母（标点符号算一个字母)
  req.SampleRate = 16000 # type: int # 音频采样率，16000：16k，8000：8k，默认16k
  # 通过client对象调用想要访问的接口，需要传入请求对象
  rep = client.TextToVoice(req)
  # rep为响应对象
  print(rep)
  """
    {
    "Audio": "UklGRlR/AABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YSx9AAD+////AQD//wAAAAAAAAIAAQADAAMABgAEAAYABQAGAAUABwAIAAgACQAAE......AAgACAAEAAgADAAIAAwACAAQAAwACAAIAAgADAAMAAgACAAIAAwABAAAAAAAAAAAAAAD/////AAAAAAAA//8AAP///v/9//7//v///////v8AAP///////wAA/////wAA/////wAAAAAAAAAAAAAAAAAAAAAAAAAA",
    "RequestId": "9a7a1615-3e09-4db2-8032-5c6f497f7e6a",
    "SessionId": "session-1234"
    }
    Audio对应的值为经过base64编码,
    RequestId为返回的唯一请求id,
    SessionId为发送请求时传入的id即uuid4()
  """
  # content为base64解码后的二进制流
  content = b64decode(rep.Audio)
  # I/O操作
  with open('voice.wav', 'wb') as f:
    f.write(content)
except TencentCloudSDKException as e:
  print(e)