Yandex Cloud
Search
Contact UsGet started
  • Pricing
  • Customer Stories
  • Documentation
  • Blog
  • All Services
  • System Status
    • Featured
    • Infrastructure & Network
    • Data Platform
    • Containers
    • Developer tools
    • Serverless
    • Security
    • Monitoring & Resources
    • AI for business
    • Business tools
  • All Solutions
    • By industry
    • By use case
    • Economics and Pricing
    • Security
    • Technical Support
    • Start testing with double trial credits
    • Cloud credits to scale your IT product
    • Gateway to Russia
    • Cloud for Startups
    • Center for Technologies and Society
    • Yandex Cloud Partner program
  • Pricing
  • Customer Stories
  • Documentation
  • Blog
© 2025 Direct Cursus Technology L.L.C.
Yandex SpeechKit Hybrid
  • System requirements
  • Service architecture
    • Service settings
    • Monitoring SpeechKit Hybrid services
    • Envoy monitoring
    • Logging levels
    • TLS termination
  • API authentication
  • Pricing policy
  • SpeechKit Hybrid releases

In this article:

  • Reserved ports
  • Run component command example
  1. SpeechKit Hybrid administration
  2. Service settings

SpeechKit Hybrid settings

Written by
Yandex Cloud
Updated at March 3, 2025
  • Reserved ports
  • Run component command example

To configure the service components, provide each setting in the respective environment variable in the --env parameter of the command to run the component's Docker container:

docker run --it \
    --env <setting_1>=<value> \
    --env <setting_2>=<value> \
    ... \
    <container_name>

If you run containers using the docker compose command, add or edit the service's description settings in the environment section of the docker-compose.yaml file.

Component Environment variable Setting description
All LOGGING_LEVEL Logging level. The default value is INFO.
STT/TTS server SERVICE_PORT Port for requests to speech/text processing services
STT server LICENSE_SERVICE_ENDPOINTS FQDN License server and the service registration port specified in the UPSTREAM_ASR_REGISTRATIONS_SERVER_PORT setting in <FQDN_License_server>:<port> format.
TTS server LICENSE_SERVICE_ENDPOINTS FQDN License server and the service registration port specified in the UPSTREAM_TTS_REGISTRATIONS_SERVER_PORT setting in <FQDN_License_server>:<port> format.
License server POSTGRES_JDBC_URL URL for connecting to the DB server of the PostgreSQL consumption metric storage in jdbc:postgresql://host:<port>/database?properties format.
License server POSTGRES_PASSWORD Password to access the DB server of the consumption metric storage.
License server POSTGRES_USER Username for the DB server of the consumption metric storage.
License server PROMETHEUS_PORT Prometheus port for sending the service's metrics. The default value is 8003.
License server STATIC_API_KEY API key for authentication on the gateway for incoming consumption metrics in the Cloud Billing licensing model.
License server UPSTREAM_ASR_PROXY_PORT Port for handling incoming requests for the STT server. Used by Envoy. The default value is 8080.
License server UPSTREAM_ASR_REGISTRATIONS_SERVER_PORT Speech recognition service registration port. The default value is 8087.
License server UPSTREAM_TTS_PROXY_PORT Port for handling incoming requests for the TTS server. Used by Envoy. The default value is 9080.
License server UPSTREAM_TTS_REGISTRATIONS_SERVER_PORT Speech synthesis service registration port. The default value is 9087.

Reserved portsReserved ports

There are ports in the service that you cannot configure:

  • License server: Port for responses from Envoy, 8086.
  • STT/TTS server: Monitoring port, 17002.

Run component command exampleRun component command example

The example assumes that the license server runs on host 172.10.19.12 with port 8083 to register an STT application. To run the STT server, use this command:

docker run -it \
   --env LICENSE_SERVICE_ENDPOINTS=172.10.19.12:8083 \
   --env SERVICE_PORT=17019 \
   cr.yandex/<registry_ID>/stt/cpu_x86_64/stt_server:<STT_application_version>

Where:

  • <registry_ID>: Registry with Docker images for deployment of SpeechKit Hybrid components.
  • <STT_application_version>: Provided STT version.

Was the article helpful?

Previous
Performance testing
Next
Monitoring SpeechKit Hybrid services
© 2025 Direct Cursus Technology L.L.C.