Get Recognition Results
AI-generatedSummary
Poll for and retrieve speech recognition results from a previously started asynchronous Yandex Cloud SpeechKit STT operation by its Operation ID, optionally auto-polling until completion or timeout.
Inputs
- Operation ID (required) — The Operation ID string returned from the Recognize Audio operation to identify the asynchronous recognition task.
- Polling Options — Settings to control polling behavior: 'Poll Interval (Seconds)' sets the wait time between attempts (default 5s), 'Max Attempts' limits the number of polling tries before timeout (default 60), and 'Return Partial Results' dictates whether to output partial transcription data if final results are not yet ready.
Output shape
A single JSON object containing recognition status and results.
Returns a JSON with 'success', 'operationId', 'status' indicating RUNNING or DONE, 'text' containing the final recognized text if done, 'finalResults' and 'partialResults' arrays, channelTag if available, and attemptsUsed. If the recognition operation times out without completion and partial results are enabled, partial transcription text is returned with status RUNNING; otherwise, an error is thrown.
Examples
Example 1: Retrieve completed transcription text for a previously started STT operation.
Provide the Operation ID from a previous Recognize Audio action, set poll interval and max attempts as needed, and disable partial results to wait for final transcription.