Speech service in sovereign clouds

Microsoft Azure operated by 21Vianet

Available to organizations with a business presence in China. See more information about Microsoft Azure operated by 21Vianet here.

• Azure portal:
  • https://portal.azure.cn/
• Regions:
  • China East 2
  • China North 2
  • China North 3
• Available pricing tiers:
  • Free (F0) and Standard (S0). See more details here
• Supported features:
  • Speech to text
    • Custom speech (Acoustic Model (AM) and Language Model (LM) adaptation)
      • Speech Studio
    • Pronunciation assessment
  • Text to speech
    • Standard voice
    • Neural voice
  • Speech translator
• Unsupported features:
  • Custom commands
  • Custom neural voice
  • Personal voice
  • Text to speech avatar
• Supported languages:
  • See the list of supported languages here

Endpoint information

This section contains Speech service endpoint information for the usage with Speech SDK, Speech to text REST API, and Text to speech REST API.

Speech service REST API

Speech service REST API endpoints in Microsoft Azure operated by 21Vianet have the following format:

Replace <REGION_IDENTIFIER> with the identifier matching the region of your subscription from this table:

Speech SDK

For Speech SDK in sovereign clouds you need to use "from host / with host" instantiation of SpeechConfig class or --host option of Speech CLI. (You can also use "from endpoint / with endpoint" instantiation and --endpoint Speech CLI option).

SpeechConfig class should be instantiated like this:

var config = SpeechConfig.FromHost("azCnHost", subscriptionKey);

auto config = SpeechConfig::FromHost("azCnHost", subscriptionKey);

SpeechConfig config = SpeechConfig.fromHost("azCnHost", subscriptionKey);

import azure.cognitiveservices.speech as speechsdk
speech_config = speechsdk.SpeechConfig(host="azCnHost", subscription=subscriptionKey)

SPXSpeechConfiguration *speechConfig = [[SPXSpeechConfiguration alloc] initWithHost:"azCnHost" subscription:subscriptionKey];

Speech CLI should be used like this (note the --host option):

spx recognize --host "azCnHost" --file myaudio.wav

Replace subscriptionKey with your Speech resource key. Replace azCnHost with the expression matching the required service offering and the region of your subscription from this table: