The top-level message returned from the
StreamingDetectIntent
method.
Multiple response messages can be returned in order:
If the input was set to streaming audio, the first one or more messages contain
recognition_result
. Eachrecognition_result
represents a more complete transcript of what the user said. The lastrecognition_result
hasis_final
set totrue
.The next message contains
response_id
,query_result
,alternative_query_results
and optionallywebhook_status
if a WebHook was called.If
output_audio_config
was specified in the request or agent-level speech synthesizer is configured, all subsequent messages containoutput_audio
andoutput_audio_config
.The result of speech recognition.
If Knowledge Connectors are enabled, there could be more than one result returned for a given query or event, and this field will contain all results except for the top one, which is captured in query_result. The alternative results are ordered by decreasing
QueryResult.intent_detection_confidence
. If Knowledge Connectors are disabled, this field will be empty until multiple responses for regular intents are supported, at which point those additional results will be surfaced here.The audio data bytes encoded as specified in the request. Note: The output audio is generated based on the values of default platform text responses found in the
query_result.fulfillment_messages
field. If multiple default text responses exist, they will be concatenated when generating audio. If no default platform text responses exist, the generated audio content will be empty.