Speech To Text Capabilities

List supported STT providers

Retrieve the canonical list of supported speech-to-text providers, models, accepted language codes, and the service types each model supports.

Service types:

  • streaming — standalone WebSocket transcription via /speech-to-text/transcription.
  • file_based — file-based transcription via /ai/audio/transcriptions.
  • in_call — live call transcription via Call Control transcription_start.
  • ai_assistant — STT configured on a Call Control AI Assistant via voice-assistant TranscriptionConfig (covers both live-streaming and non-streaming/batch models).

Use this endpoint to discover which (provider, model) combinations are available for the surface you need, and which language codes each accepts. auto in a languages array indicates the provider performs language detection.

get/speech-to-text/providers

Query parameters

service_type'streaming' | 'file_based' | 'in_call' | 'ai_assistant'

Service surface a model is available on. ai_assistant is the STT surface configured via Call Control voice-assistant transcription; it covers both live-streaming and non-streaming/batch models (matching the TranscriptionConfig.model enum on call-control voice assistants).

Filter to entries that support the given service type. For backward compatibility with the values that briefly shipped before the product-aligned rename, the legacy aliases file_transcription, in_call_transcription, and ai_assistant_transcription are silently accepted and normalized to file_based, in_call, and ai_assistant respectively. The response always emits the canonical (post-rename) values.

provider'deepgram' | 'speechmatics' | 'assemblyai' | 'xai' | 'soniox' | 'azure' | 'openai' | 'google' | 'telnyx'

Filter to entries for a specific STT provider. The enum mirrors the providers advertised across the speech-to-text spec (including google and telnyx, which are accepted as WebSocket transcription engines). A provider that has no models currently registered for any service type will return an empty data array rather than an error.

Response

List of supported STT providers and models.

Example response

{
  "data": [
    {
      "provider": "deepgram",
      "model": "deepgram/nova-3",
      "service_types": [
        {
          "languages": [
            "en",
            "es",
            "fr"
          ]
        }
      ]
    }
  ],
  "meta": {
    "total": 12
  }
}

Changes

Changed in 2 of the 47 revisions of this API.49

    • removed the enum value file_transcription from the query request parameter service_type

      request-parameter-enum-value-removed

    • removed the enum value in_call_transcription from the query request parameter service_type

      request-parameter-enum-value-removed

    • the /// response's property type changed from string to object for status

      response-property-type-changed

    • removed the required property // from the response with the status

      response-required-property-removed

    • added the new enum value ai_assistant to the query request parameter service_type

      request-parameter-enum-value-added

    • added the new enum value file_based to the query request parameter service_type

      request-parameter-enum-value-added

    • added the new enum value in_call to the query request parameter service_type

      request-parameter-enum-value-added

    • removed the file_transcription enum value from the /// response property for the response status

      response-property-enum-value-removed

    • removed the in_call_transcription enum value from the /// response property for the response status

      response-property-enum-value-removed

    • removed the streaming enum value from the /// response property for the response status

      response-property-enum-value-removed

    • added the required property //// to the response with the status

      response-required-property-added

    • added the required property //// to the response with the status

      response-required-property-added

    • endpoint added

      endpoint-added