SpeechSense

Загрузка диалогов в Yandex SpeechSense (Talk Analytics) по gRPC

Talk → Upload Audio

AI-generated

Summary

Uploads an input binary audio file to a Yandex SpeechSense connection using the TalkService gRPC API. Creates a new talk when Talk ID is empty, or includes the supplied Talk ID in the upload request.

Inputs

  • Connection ID (required) — SpeechSense connection that receives the talk.
  • Input Binary Field (required) — Name of the incoming binary property containing the audio file; defaults to data.
  • Audio Type — Audio container type: MP3 (default), WAV, or OGG / Opus.
  • Talk ID — If provided, adds this talk ID to the upload request; if empty, no talk ID is sent.
  • Metadata Fields (JSON) — Object of metadata fields. Values are converted to strings; null and undefined values are omitted. Defaults to {}.
  • Users (JSON) — Array of participants with id, role, and optional fields. Participant field values are converted to strings. Defaults to [].

Output shape

One output item per input item, with JSON containing talk_id and connection_id.

The returned talk_id is null if the API response does not include one. Each output item is paired with its corresponding input. With Continue On Fail enabled, a failed item returns { "error": "..." }; otherwise the operation raises an error.

Links

Discussion