2026/10/21 - Amazon Lex Model Building V2 - 3 updated api methods
{'generativeAISettings': {'runtimeSettings': {'nluImprovement': {'assistedNluMode': {'DialogManager'}}}}}
Creates a locale in the bot. The locale contains the intents and slot types that the bot uses in conversations with users in the specified language and locale. You must add a locale to a bot before you can add intents and slot types to the bot.
See also: AWS API Documentation
Request Syntax
client.create_bot_locale(
botId='string',
botVersion='string',
localeId='string',
description='string',
nluIntentConfidenceThreshold=123.0,
voiceSettings={
'engine': 'standard'|'neural'|'long-form'|'generative',
'voiceId': 'string'
},
unifiedSpeechSettings={
'speechFoundationModel': {
'modelArn': 'string',
'voiceId': 'string'
}
},
audioFillerSettings={
'enabled': True|False,
'audioType': 'MELODY_CHIPPER_CHIME'|'MELODY_CURIOUS_CRAWL'|'MELODY_RISING_RIPPLE'|'MELODY_PATIENT_PING'|'MELODY_PONDERING_PONG'|'TYPING_KINETIC_KEYS'|'TYPING_QUIET_QWERTY',
'startDelayInMilliseconds': 123,
'minimumPlayDurationInMilliseconds': 123,
'responseDeliveryDelayInMilliseconds': 123
},
speechRecognitionSettings={
'speechModelPreference': 'Standard'|'Neural'|'Deepgram'|'Advanced',
'speechModelConfig': {
'deepgramConfig': {
'apiTokenSecretArn': 'string',
'modelId': 'string'
}
}
},
generativeAISettings={
'runtimeSettings': {
'slotResolutionImprovement': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'nluImprovement': {
'enabled': True|False,
'assistedNluMode': 'Primary'|'Fallback'|'DialogManager',
'intentDisambiguationSettings': {
'enabled': True|False,
'maxDisambiguationIntents': 123,
'customDisambiguationMessage': 'string'
}
}
},
'buildtimeSettings': {
'descriptiveBotBuilder': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'sampleUtteranceGeneration': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
}
}
},
speechDetectionSensitivity='Default'|'HighNoiseTolerance'|'MaximumNoiseTolerance',
speakerDiarizationSettings={
'enabled': True|False
}
)
string
[REQUIRED]
The identifier of the bot to create the locale for.
string
[REQUIRED]
The version of the bot to create the locale for. This can only be the draft version of the bot.
string
[REQUIRED]
The identifier of the language and locale that the bot will be used in. The string must match one of the supported locales. All of the intents, slot types, and slots used in the bot must have the same locale. For more information, see Supported languages.
string
A description of the bot locale. Use this to help identify the bot locale in lists.
float
[REQUIRED]
Determines the threshold where Amazon Lex will insert the AMAZON.FallbackIntent, AMAZON.KendraSearchIntent, or both when returning alternative intents. AMAZON.FallbackIntent and AMAZON.KendraSearchIntent are only inserted if they are configured for the bot.
For example, suppose a bot is configured with the confidence threshold of 0.80 and the AMAZON.FallbackIntent. Amazon Lex returns three alternative intents with the following confidence scores: IntentA (0.70), IntentB (0.60), IntentC (0.50). The response from the RecognizeText operation would be:
AMAZON.FallbackIntent
IntentA
IntentB
IntentC
dict
The Amazon Polly voice ID that Amazon Lex uses for voice interaction with the user.
engine (string) --
Indicates the type of Amazon Polly voice that Amazon Lex should use for voice interaction with the user. For more information, see the engine parameter of the SynthesizeSpeech operation in the Amazon Polly developer guide.
If you do not specify a value, the default is standard.
voiceId (string) -- [REQUIRED]
The identifier of the Amazon Polly voice to use.
dict
Unified speech settings to configure for the new bot locale.
speechFoundationModel (dict) -- [REQUIRED]
The foundation model configuration to use for unified speech processing capabilities.
modelArn (string) -- [REQUIRED]
The Amazon Resource Name (ARN) of the foundation model used for speech processing.
voiceId (string) --
The identifier of the voice to use for speech synthesis with the foundation model.
dict
Audio filler settings to configure for the new bot locale. When enabled, Amazon Lex plays a brief background audio filler during speech-to-speech interactions to mask processing delays. Requires unifiedSpeechSettings (speech-to-speech) to be configured on the bot locale.
enabled (boolean) --
Specifies whether audio filler playback is enabled for the bot locale. Set to true to play filler audio while Amazon Lex processes a user utterance. Set to false to disable filler audio.
audioType (string) --
The identifier of the audio filler to play while Amazon Lex processes the user's input. This field is required when enabled is true.
startDelayInMilliseconds (integer) --
The time, in milliseconds, to wait after the end of the user's utterance before starting audio filler playback. Valid range is 500 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 2500 milliseconds.
minimumPlayDurationInMilliseconds (integer) --
The minimum time, in milliseconds, that audio filler plays once it has started, even if the bot response becomes ready sooner. Valid range is 1000 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 3000 milliseconds.
responseDeliveryDelayInMilliseconds (integer) --
The silent delay, in milliseconds, inserted between the end of audio filler playback and the start of the bot's response. Valid range is 200 to 1000 milliseconds. If not specified, Amazon Lex uses a default of 500 milliseconds.
dict
Speech-to-text settings to configure for the new bot locale.
speechModelPreference (string) --
The speech-to-text model to use.
speechModelConfig (dict) --
Configuration settings for the selected speech-to-text model.
deepgramConfig (dict) --
Configuration settings for using Deepgram as the speech-to-text provider.
apiTokenSecretArn (string) -- [REQUIRED]
The Amazon Resource Name (ARN) of the Secrets Manager secret that contains the Deepgram API token.
modelId (string) --
The identifier of the Deepgram speech-to-text model to use for processing speech input.
dict
Contains specifications about the generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
runtimeSettings (dict) --
Contains specifications about the Amazon Lex runtime generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
slotResolutionImprovement (dict) --
An object containing specifications for the assisted slot resolution feature.
enabled (boolean) -- [REQUIRED]
Specifies whether assisted slot resolution is turned on or off.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to assist slot resolution.
modelArn (string) -- [REQUIRED]
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) -- [REQUIRED]
The unique guardrail id for the Bedrock guardrail configuration.
version (string) -- [REQUIRED]
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
nluImprovement (dict) --
An object containing specifications for the Assisted NLU feature within the bot's runtime settings. These settings determine how the bot processes and interprets user utterances during conversations.
enabled (boolean) -- [REQUIRED]
Determines whether the Assisted NLU feature is enabled for the bot. When set to true, Amazon Lex uses advanced models to improve intent recognition and slot resolution, with the default being false.
assistedNluMode (string) --
Specifies the mode for Assisted NLU operation. Use Primary to make Assisted NLU the primary intent recognition method, Fallback to use it only when standard NLU confidence is low, or DialogManager to use a large language model to manage the conversation, including intent recognition, slot elicitation, and confirmation.
intentDisambiguationSettings (dict) --
An object containing specifications for the Intent Disambiguation feature within the Assisted NLU settings. These settings determine how the bot handles ambiguous user inputs that could match multiple intents.
enabled (boolean) -- [REQUIRED]
Determines whether the Intent Disambiguation feature is enabled. When set to true, Amazon Lex will present disambiguation options to users when multiple intents could match their input, with the default being false.
maxDisambiguationIntents (integer) --
Specifies the maximum number of intent options (2-5) to present to users when disambiguation is needed. This setting determines how many intent options will be shown to users when the system detects ambiguous input. The default value is 3.
customDisambiguationMessage (string) --
Provides a custom message that will be displayed before presenting the disambiguation options to users. This message helps set the context for users and can be customized to match your bot's tone and brand. If not specified, a default message will be used.
buildtimeSettings (dict) --
Contains specifications about the Amazon Lex build time generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
descriptiveBotBuilder (dict) --
An object containing specifications for the descriptive bot building feature.
enabled (boolean) -- [REQUIRED]
Specifies whether the descriptive bot building feature is activated or not.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) -- [REQUIRED]
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) -- [REQUIRED]
The unique guardrail id for the Bedrock guardrail configuration.
version (string) -- [REQUIRED]
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
sampleUtteranceGeneration (dict) --
Contains specifications for the sample utterance generation feature.
enabled (boolean) -- [REQUIRED]
Specifies whether to enable sample utterance generation or not.
bedrockModelSpecification (dict) --
Contains information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) -- [REQUIRED]
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) -- [REQUIRED]
The unique guardrail id for the Bedrock guardrail configuration.
version (string) -- [REQUIRED]
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
string
The sensitivity level for voice activity detection (VAD) in the bot locale. This setting helps optimize speech recognition accuracy by adjusting how the system responds to background noise during voice interactions.
dict
The speaker diarization settings to configure for the new bot locale. When enabled, Amazon Lex restricts speech detection to the primary (loudest) speaker during streaming audio conversations.
enabled (boolean) -- [REQUIRED]
Specifies whether speaker diarization is enabled for the bot locale. Set to true to have Amazon Lex treat speech from speakers other than the primary speaker as non-speech. Set to false to disable speaker diarization and rely on voice activity detection alone.
dict
Response Syntax
{
'botId': 'string',
'botVersion': 'string',
'localeName': 'string',
'localeId': 'string',
'description': 'string',
'nluIntentConfidenceThreshold': 123.0,
'voiceSettings': {
'engine': 'standard'|'neural'|'long-form'|'generative',
'voiceId': 'string'
},
'unifiedSpeechSettings': {
'speechFoundationModel': {
'modelArn': 'string',
'voiceId': 'string'
}
},
'audioFillerSettings': {
'enabled': True|False,
'audioType': 'MELODY_CHIPPER_CHIME'|'MELODY_CURIOUS_CRAWL'|'MELODY_RISING_RIPPLE'|'MELODY_PATIENT_PING'|'MELODY_PONDERING_PONG'|'TYPING_KINETIC_KEYS'|'TYPING_QUIET_QWERTY',
'startDelayInMilliseconds': 123,
'minimumPlayDurationInMilliseconds': 123,
'responseDeliveryDelayInMilliseconds': 123
},
'speechRecognitionSettings': {
'speechModelPreference': 'Standard'|'Neural'|'Deepgram'|'Advanced',
'speechModelConfig': {
'deepgramConfig': {
'apiTokenSecretArn': 'string',
'modelId': 'string'
}
}
},
'botLocaleStatus': 'Creating'|'Building'|'Built'|'ReadyExpressTesting'|'Failed'|'Deleting'|'NotBuilt'|'Importing'|'Processing',
'creationDateTime': datetime(2015, 1, 1),
'generativeAISettings': {
'runtimeSettings': {
'slotResolutionImprovement': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'nluImprovement': {
'enabled': True|False,
'assistedNluMode': 'Primary'|'Fallback'|'DialogManager',
'intentDisambiguationSettings': {
'enabled': True|False,
'maxDisambiguationIntents': 123,
'customDisambiguationMessage': 'string'
}
}
},
'buildtimeSettings': {
'descriptiveBotBuilder': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'sampleUtteranceGeneration': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
}
}
},
'speechDetectionSensitivity': 'Default'|'HighNoiseTolerance'|'MaximumNoiseTolerance',
'speakerDiarizationSettings': {
'enabled': True|False
}
}
Response Structure
(dict) --
botId (string) --
The specified bot identifier.
botVersion (string) --
The specified bot version.
localeName (string) --
The specified locale name.
localeId (string) --
The specified locale identifier.
description (string) --
The specified description of the bot locale.
nluIntentConfidenceThreshold (float) --
The specified confidence threshold for inserting the AMAZON.FallbackIntent and AMAZON.KendraSearchIntent intents.
voiceSettings (dict) --
The Amazon Polly voice ID that Amazon Lex uses for voice interaction with the user.
engine (string) --
Indicates the type of Amazon Polly voice that Amazon Lex should use for voice interaction with the user. For more information, see the engine parameter of the SynthesizeSpeech operation in the Amazon Polly developer guide.
If you do not specify a value, the default is standard.
voiceId (string) --
The identifier of the Amazon Polly voice to use.
unifiedSpeechSettings (dict) --
The unified speech settings configured for the created bot locale.
speechFoundationModel (dict) --
The foundation model configuration to use for unified speech processing capabilities.
modelArn (string) --
The Amazon Resource Name (ARN) of the foundation model used for speech processing.
voiceId (string) --
The identifier of the voice to use for speech synthesis with the foundation model.
audioFillerSettings (dict) --
The audio filler settings configured for the created bot locale.
enabled (boolean) --
Specifies whether audio filler playback is enabled for the bot locale. Set to true to play filler audio while Amazon Lex processes a user utterance. Set to false to disable filler audio.
audioType (string) --
The identifier of the audio filler to play while Amazon Lex processes the user's input. This field is required when enabled is true.
startDelayInMilliseconds (integer) --
The time, in milliseconds, to wait after the end of the user's utterance before starting audio filler playback. Valid range is 500 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 2500 milliseconds.
minimumPlayDurationInMilliseconds (integer) --
The minimum time, in milliseconds, that audio filler plays once it has started, even if the bot response becomes ready sooner. Valid range is 1000 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 3000 milliseconds.
responseDeliveryDelayInMilliseconds (integer) --
The silent delay, in milliseconds, inserted between the end of audio filler playback and the start of the bot's response. Valid range is 200 to 1000 milliseconds. If not specified, Amazon Lex uses a default of 500 milliseconds.
speechRecognitionSettings (dict) --
The speech-to-text settings configured for the created bot locale.
speechModelPreference (string) --
The speech-to-text model to use.
speechModelConfig (dict) --
Configuration settings for the selected speech-to-text model.
deepgramConfig (dict) --
Configuration settings for using Deepgram as the speech-to-text provider.
apiTokenSecretArn (string) --
The Amazon Resource Name (ARN) of the Secrets Manager secret that contains the Deepgram API token.
modelId (string) --
The identifier of the Deepgram speech-to-text model to use for processing speech input.
botLocaleStatus (string) --
The status of the bot.
When the status is Creating the bot locale is being configured. When the status is Building Amazon Lex is building the bot for testing and use.
If the status of the bot is ReadyExpressTesting, you can test the bot using the exact utterances specified in the bots' intents. When the bot is ready for full testing or to run, the status is Built.
If there was a problem with building the bot, the status is Failed. If the bot was saved but not built, the status is NotBuilt.
creationDateTime (datetime) --
A timestamp specifying the date and time that the bot locale was created.
generativeAISettings (dict) --
Contains specifications about the generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
runtimeSettings (dict) --
Contains specifications about the Amazon Lex runtime generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
slotResolutionImprovement (dict) --
An object containing specifications for the assisted slot resolution feature.
enabled (boolean) --
Specifies whether assisted slot resolution is turned on or off.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to assist slot resolution.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
nluImprovement (dict) --
An object containing specifications for the Assisted NLU feature within the bot's runtime settings. These settings determine how the bot processes and interprets user utterances during conversations.
enabled (boolean) --
Determines whether the Assisted NLU feature is enabled for the bot. When set to true, Amazon Lex uses advanced models to improve intent recognition and slot resolution, with the default being false.
assistedNluMode (string) --
Specifies the mode for Assisted NLU operation. Use Primary to make Assisted NLU the primary intent recognition method, Fallback to use it only when standard NLU confidence is low, or DialogManager to use a large language model to manage the conversation, including intent recognition, slot elicitation, and confirmation.
intentDisambiguationSettings (dict) --
An object containing specifications for the Intent Disambiguation feature within the Assisted NLU settings. These settings determine how the bot handles ambiguous user inputs that could match multiple intents.
enabled (boolean) --
Determines whether the Intent Disambiguation feature is enabled. When set to true, Amazon Lex will present disambiguation options to users when multiple intents could match their input, with the default being false.
maxDisambiguationIntents (integer) --
Specifies the maximum number of intent options (2-5) to present to users when disambiguation is needed. This setting determines how many intent options will be shown to users when the system detects ambiguous input. The default value is 3.
customDisambiguationMessage (string) --
Provides a custom message that will be displayed before presenting the disambiguation options to users. This message helps set the context for users and can be customized to match your bot's tone and brand. If not specified, a default message will be used.
buildtimeSettings (dict) --
Contains specifications about the Amazon Lex build time generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
descriptiveBotBuilder (dict) --
An object containing specifications for the descriptive bot building feature.
enabled (boolean) --
Specifies whether the descriptive bot building feature is activated or not.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
sampleUtteranceGeneration (dict) --
Contains specifications for the sample utterance generation feature.
enabled (boolean) --
Specifies whether to enable sample utterance generation or not.
bedrockModelSpecification (dict) --
Contains information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
speechDetectionSensitivity (string) --
The sensitivity level for voice activity detection (VAD) that was specified for the bot locale.
speakerDiarizationSettings (dict) --
The speaker diarization settings configured for the created bot locale.
enabled (boolean) --
Specifies whether speaker diarization is enabled for the bot locale. Set to true to have Amazon Lex treat speech from speakers other than the primary speaker as non-speech. Set to false to disable speaker diarization and rely on voice activity detection alone.
{'generativeAISettings': {'runtimeSettings': {'nluImprovement': {'assistedNluMode': {'DialogManager'}}}}}
Describes the settings that a bot has for a specific locale.
See also: AWS API Documentation
Request Syntax
client.describe_bot_locale(
botId='string',
botVersion='string',
localeId='string'
)
string
[REQUIRED]
The identifier of the bot associated with the locale.
string
[REQUIRED]
The version of the bot associated with the locale.
string
[REQUIRED]
The unique identifier of the locale to describe. The string must match one of the supported locales. For more information, see Supported languages.
dict
Response Syntax
{
'botId': 'string',
'botVersion': 'string',
'localeId': 'string',
'localeName': 'string',
'description': 'string',
'nluIntentConfidenceThreshold': 123.0,
'voiceSettings': {
'engine': 'standard'|'neural'|'long-form'|'generative',
'voiceId': 'string'
},
'unifiedSpeechSettings': {
'speechFoundationModel': {
'modelArn': 'string',
'voiceId': 'string'
}
},
'audioFillerSettings': {
'enabled': True|False,
'audioType': 'MELODY_CHIPPER_CHIME'|'MELODY_CURIOUS_CRAWL'|'MELODY_RISING_RIPPLE'|'MELODY_PATIENT_PING'|'MELODY_PONDERING_PONG'|'TYPING_KINETIC_KEYS'|'TYPING_QUIET_QWERTY',
'startDelayInMilliseconds': 123,
'minimumPlayDurationInMilliseconds': 123,
'responseDeliveryDelayInMilliseconds': 123
},
'speechRecognitionSettings': {
'speechModelPreference': 'Standard'|'Neural'|'Deepgram'|'Advanced',
'speechModelConfig': {
'deepgramConfig': {
'apiTokenSecretArn': 'string',
'modelId': 'string'
}
}
},
'intentsCount': 123,
'slotTypesCount': 123,
'botLocaleStatus': 'Creating'|'Building'|'Built'|'ReadyExpressTesting'|'Failed'|'Deleting'|'NotBuilt'|'Importing'|'Processing',
'failureReasons': [
'string',
],
'creationDateTime': datetime(2015, 1, 1),
'lastUpdatedDateTime': datetime(2015, 1, 1),
'lastBuildSubmittedDateTime': datetime(2015, 1, 1),
'botLocaleHistoryEvents': [
{
'event': 'string',
'eventDate': datetime(2015, 1, 1)
},
],
'recommendedActions': [
'string',
],
'generativeAISettings': {
'runtimeSettings': {
'slotResolutionImprovement': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'nluImprovement': {
'enabled': True|False,
'assistedNluMode': 'Primary'|'Fallback'|'DialogManager',
'intentDisambiguationSettings': {
'enabled': True|False,
'maxDisambiguationIntents': 123,
'customDisambiguationMessage': 'string'
}
}
},
'buildtimeSettings': {
'descriptiveBotBuilder': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'sampleUtteranceGeneration': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
}
}
},
'speechDetectionSensitivity': 'Default'|'HighNoiseTolerance'|'MaximumNoiseTolerance',
'speakerDiarizationSettings': {
'enabled': True|False
}
}
Response Structure
(dict) --
botId (string) --
The identifier of the bot associated with the locale.
botVersion (string) --
The version of the bot associated with the locale.
localeId (string) --
The unique identifier of the described locale.
localeName (string) --
The name of the locale.
description (string) --
The description of the locale.
nluIntentConfidenceThreshold (float) --
The confidence threshold where Amazon Lex inserts the AMAZON.FallbackIntent and AMAZON.KendraSearchIntent intents in the list of possible intents for an utterance.
voiceSettings (dict) --
The Amazon Polly voice Amazon Lex uses for voice interaction with the user.
engine (string) --
Indicates the type of Amazon Polly voice that Amazon Lex should use for voice interaction with the user. For more information, see the engine parameter of the SynthesizeSpeech operation in the Amazon Polly developer guide.
If you do not specify a value, the default is standard.
voiceId (string) --
The identifier of the Amazon Polly voice to use.
unifiedSpeechSettings (dict) --
The unified speech settings configured for the bot locale.
speechFoundationModel (dict) --
The foundation model configuration to use for unified speech processing capabilities.
modelArn (string) --
The Amazon Resource Name (ARN) of the foundation model used for speech processing.
voiceId (string) --
The identifier of the voice to use for speech synthesis with the foundation model.
audioFillerSettings (dict) --
The audio filler settings configured for the bot locale.
enabled (boolean) --
Specifies whether audio filler playback is enabled for the bot locale. Set to true to play filler audio while Amazon Lex processes a user utterance. Set to false to disable filler audio.
audioType (string) --
The identifier of the audio filler to play while Amazon Lex processes the user's input. This field is required when enabled is true.
startDelayInMilliseconds (integer) --
The time, in milliseconds, to wait after the end of the user's utterance before starting audio filler playback. Valid range is 500 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 2500 milliseconds.
minimumPlayDurationInMilliseconds (integer) --
The minimum time, in milliseconds, that audio filler plays once it has started, even if the bot response becomes ready sooner. Valid range is 1000 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 3000 milliseconds.
responseDeliveryDelayInMilliseconds (integer) --
The silent delay, in milliseconds, inserted between the end of audio filler playback and the start of the bot's response. Valid range is 200 to 1000 milliseconds. If not specified, Amazon Lex uses a default of 500 milliseconds.
speechRecognitionSettings (dict) --
The speech-to-text settings configured for the bot locale.
speechModelPreference (string) --
The speech-to-text model to use.
speechModelConfig (dict) --
Configuration settings for the selected speech-to-text model.
deepgramConfig (dict) --
Configuration settings for using Deepgram as the speech-to-text provider.
apiTokenSecretArn (string) --
The Amazon Resource Name (ARN) of the Secrets Manager secret that contains the Deepgram API token.
modelId (string) --
The identifier of the Deepgram speech-to-text model to use for processing speech input.
intentsCount (integer) --
The number of intents defined for the locale.
slotTypesCount (integer) --
The number of slot types defined for the locale.
botLocaleStatus (string) --
The status of the bot. If the status is Failed, the reasons for the failure are listed in the failureReasons field.
failureReasons (list) --
if botLocaleStatus is Failed, Amazon Lex explains why it failed to build the bot.
(string) --
creationDateTime (datetime) --
The date and time that the locale was created.
lastUpdatedDateTime (datetime) --
The date and time that the locale was last updated.
lastBuildSubmittedDateTime (datetime) --
The date and time that the locale was last submitted for building.
botLocaleHistoryEvents (list) --
History of changes, such as when a locale is used in an alias, that have taken place for the locale.
(dict) --
Provides information about an event that occurred affecting the bot locale.
event (string) --
A description of the event that occurred.
eventDate (datetime) --
A timestamp of the date and time that the event occurred.
recommendedActions (list) --
Recommended actions to take to resolve an error in the failureReasons field.
(string) --
generativeAISettings (dict) --
Contains settings for Amazon Bedrock's generative AI features for your bot locale.
runtimeSettings (dict) --
Contains specifications about the Amazon Lex runtime generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
slotResolutionImprovement (dict) --
An object containing specifications for the assisted slot resolution feature.
enabled (boolean) --
Specifies whether assisted slot resolution is turned on or off.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to assist slot resolution.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
nluImprovement (dict) --
An object containing specifications for the Assisted NLU feature within the bot's runtime settings. These settings determine how the bot processes and interprets user utterances during conversations.
enabled (boolean) --
Determines whether the Assisted NLU feature is enabled for the bot. When set to true, Amazon Lex uses advanced models to improve intent recognition and slot resolution, with the default being false.
assistedNluMode (string) --
Specifies the mode for Assisted NLU operation. Use Primary to make Assisted NLU the primary intent recognition method, Fallback to use it only when standard NLU confidence is low, or DialogManager to use a large language model to manage the conversation, including intent recognition, slot elicitation, and confirmation.
intentDisambiguationSettings (dict) --
An object containing specifications for the Intent Disambiguation feature within the Assisted NLU settings. These settings determine how the bot handles ambiguous user inputs that could match multiple intents.
enabled (boolean) --
Determines whether the Intent Disambiguation feature is enabled. When set to true, Amazon Lex will present disambiguation options to users when multiple intents could match their input, with the default being false.
maxDisambiguationIntents (integer) --
Specifies the maximum number of intent options (2-5) to present to users when disambiguation is needed. This setting determines how many intent options will be shown to users when the system detects ambiguous input. The default value is 3.
customDisambiguationMessage (string) --
Provides a custom message that will be displayed before presenting the disambiguation options to users. This message helps set the context for users and can be customized to match your bot's tone and brand. If not specified, a default message will be used.
buildtimeSettings (dict) --
Contains specifications about the Amazon Lex build time generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
descriptiveBotBuilder (dict) --
An object containing specifications for the descriptive bot building feature.
enabled (boolean) --
Specifies whether the descriptive bot building feature is activated or not.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
sampleUtteranceGeneration (dict) --
Contains specifications for the sample utterance generation feature.
enabled (boolean) --
Specifies whether to enable sample utterance generation or not.
bedrockModelSpecification (dict) --
Contains information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
speechDetectionSensitivity (string) --
The sensitivity level for voice activity detection (VAD) configured for the bot locale.
speakerDiarizationSettings (dict) --
The speaker diarization settings configured for the bot locale.
enabled (boolean) --
Specifies whether speaker diarization is enabled for the bot locale. Set to true to have Amazon Lex treat speech from speakers other than the primary speaker as non-speech. Set to false to disable speaker diarization and rely on voice activity detection alone.
{'generativeAISettings': {'runtimeSettings': {'nluImprovement': {'assistedNluMode': {'DialogManager'}}}}}
Updates the settings that a bot has for a specific locale.
See also: AWS API Documentation
Request Syntax
client.update_bot_locale(
botId='string',
botVersion='string',
localeId='string',
description='string',
nluIntentConfidenceThreshold=123.0,
voiceSettings={
'engine': 'standard'|'neural'|'long-form'|'generative',
'voiceId': 'string'
},
unifiedSpeechSettings={
'speechFoundationModel': {
'modelArn': 'string',
'voiceId': 'string'
}
},
audioFillerSettings={
'enabled': True|False,
'audioType': 'MELODY_CHIPPER_CHIME'|'MELODY_CURIOUS_CRAWL'|'MELODY_RISING_RIPPLE'|'MELODY_PATIENT_PING'|'MELODY_PONDERING_PONG'|'TYPING_KINETIC_KEYS'|'TYPING_QUIET_QWERTY',
'startDelayInMilliseconds': 123,
'minimumPlayDurationInMilliseconds': 123,
'responseDeliveryDelayInMilliseconds': 123
},
speechRecognitionSettings={
'speechModelPreference': 'Standard'|'Neural'|'Deepgram'|'Advanced',
'speechModelConfig': {
'deepgramConfig': {
'apiTokenSecretArn': 'string',
'modelId': 'string'
}
}
},
generativeAISettings={
'runtimeSettings': {
'slotResolutionImprovement': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'nluImprovement': {
'enabled': True|False,
'assistedNluMode': 'Primary'|'Fallback'|'DialogManager',
'intentDisambiguationSettings': {
'enabled': True|False,
'maxDisambiguationIntents': 123,
'customDisambiguationMessage': 'string'
}
}
},
'buildtimeSettings': {
'descriptiveBotBuilder': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'sampleUtteranceGeneration': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
}
}
},
speechDetectionSensitivity='Default'|'HighNoiseTolerance'|'MaximumNoiseTolerance',
speakerDiarizationSettings={
'enabled': True|False
}
)
string
[REQUIRED]
The unique identifier of the bot that contains the locale.
string
[REQUIRED]
The version of the bot that contains the locale to be updated. The version can only be the DRAFT version.
string
[REQUIRED]
The identifier of the language and locale to update. The string must match one of the supported locales. For more information, see Supported languages.
string
The new description of the locale.
float
[REQUIRED]
The new confidence threshold where Amazon Lex inserts the AMAZON.FallbackIntent and AMAZON.KendraSearchIntent intents in the list of possible intents for an utterance.
dict
The new Amazon Polly voice Amazon Lex should use for voice interaction with the user.
engine (string) --
Indicates the type of Amazon Polly voice that Amazon Lex should use for voice interaction with the user. For more information, see the engine parameter of the SynthesizeSpeech operation in the Amazon Polly developer guide.
If you do not specify a value, the default is standard.
voiceId (string) -- [REQUIRED]
The identifier of the Amazon Polly voice to use.
dict
Updated unified speech settings to apply to the bot locale.
speechFoundationModel (dict) -- [REQUIRED]
The foundation model configuration to use for unified speech processing capabilities.
modelArn (string) -- [REQUIRED]
The Amazon Resource Name (ARN) of the foundation model used for speech processing.
voiceId (string) --
The identifier of the voice to use for speech synthesis with the foundation model.
dict
Updated audio filler settings to apply to the bot locale. When enabled, requires unifiedSpeechSettings (speech-to-speech) to be configured on the bot locale.
enabled (boolean) --
Specifies whether audio filler playback is enabled for the bot locale. Set to true to play filler audio while Amazon Lex processes a user utterance. Set to false to disable filler audio.
audioType (string) --
The identifier of the audio filler to play while Amazon Lex processes the user's input. This field is required when enabled is true.
startDelayInMilliseconds (integer) --
The time, in milliseconds, to wait after the end of the user's utterance before starting audio filler playback. Valid range is 500 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 2500 milliseconds.
minimumPlayDurationInMilliseconds (integer) --
The minimum time, in milliseconds, that audio filler plays once it has started, even if the bot response becomes ready sooner. Valid range is 1000 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 3000 milliseconds.
responseDeliveryDelayInMilliseconds (integer) --
The silent delay, in milliseconds, inserted between the end of audio filler playback and the start of the bot's response. Valid range is 200 to 1000 milliseconds. If not specified, Amazon Lex uses a default of 500 milliseconds.
dict
Updated speech-to-text settings to apply to the bot locale.
speechModelPreference (string) --
The speech-to-text model to use.
speechModelConfig (dict) --
Configuration settings for the selected speech-to-text model.
deepgramConfig (dict) --
Configuration settings for using Deepgram as the speech-to-text provider.
apiTokenSecretArn (string) -- [REQUIRED]
The Amazon Resource Name (ARN) of the Secrets Manager secret that contains the Deepgram API token.
modelId (string) --
The identifier of the Deepgram speech-to-text model to use for processing speech input.
dict
Contains settings for generative AI features powered by Amazon Bedrock for your bot locale. Use this object to turn generative AI features on and off. Pricing may differ if you turn a feature on. For more information, see LINK.
runtimeSettings (dict) --
Contains specifications about the Amazon Lex runtime generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
slotResolutionImprovement (dict) --
An object containing specifications for the assisted slot resolution feature.
enabled (boolean) -- [REQUIRED]
Specifies whether assisted slot resolution is turned on or off.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to assist slot resolution.
modelArn (string) -- [REQUIRED]
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) -- [REQUIRED]
The unique guardrail id for the Bedrock guardrail configuration.
version (string) -- [REQUIRED]
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
nluImprovement (dict) --
An object containing specifications for the Assisted NLU feature within the bot's runtime settings. These settings determine how the bot processes and interprets user utterances during conversations.
enabled (boolean) -- [REQUIRED]
Determines whether the Assisted NLU feature is enabled for the bot. When set to true, Amazon Lex uses advanced models to improve intent recognition and slot resolution, with the default being false.
assistedNluMode (string) --
Specifies the mode for Assisted NLU operation. Use Primary to make Assisted NLU the primary intent recognition method, Fallback to use it only when standard NLU confidence is low, or DialogManager to use a large language model to manage the conversation, including intent recognition, slot elicitation, and confirmation.
intentDisambiguationSettings (dict) --
An object containing specifications for the Intent Disambiguation feature within the Assisted NLU settings. These settings determine how the bot handles ambiguous user inputs that could match multiple intents.
enabled (boolean) -- [REQUIRED]
Determines whether the Intent Disambiguation feature is enabled. When set to true, Amazon Lex will present disambiguation options to users when multiple intents could match their input, with the default being false.
maxDisambiguationIntents (integer) --
Specifies the maximum number of intent options (2-5) to present to users when disambiguation is needed. This setting determines how many intent options will be shown to users when the system detects ambiguous input. The default value is 3.
customDisambiguationMessage (string) --
Provides a custom message that will be displayed before presenting the disambiguation options to users. This message helps set the context for users and can be customized to match your bot's tone and brand. If not specified, a default message will be used.
buildtimeSettings (dict) --
Contains specifications about the Amazon Lex build time generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
descriptiveBotBuilder (dict) --
An object containing specifications for the descriptive bot building feature.
enabled (boolean) -- [REQUIRED]
Specifies whether the descriptive bot building feature is activated or not.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) -- [REQUIRED]
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) -- [REQUIRED]
The unique guardrail id for the Bedrock guardrail configuration.
version (string) -- [REQUIRED]
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
sampleUtteranceGeneration (dict) --
Contains specifications for the sample utterance generation feature.
enabled (boolean) -- [REQUIRED]
Specifies whether to enable sample utterance generation or not.
bedrockModelSpecification (dict) --
Contains information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) -- [REQUIRED]
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) -- [REQUIRED]
The unique guardrail id for the Bedrock guardrail configuration.
version (string) -- [REQUIRED]
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
string
The new sensitivity level for voice activity detection (VAD) in the bot locale. This setting helps optimize speech recognition accuracy by adjusting how the system responds to background noise during voice interactions.
dict
The updated speaker diarization settings to apply to the bot locale. If you omit this field, Amazon Lex keeps the setting currently stored on the bot locale. To turn speaker diarization off, set enabled to false explicitly.
enabled (boolean) -- [REQUIRED]
Specifies whether speaker diarization is enabled for the bot locale. Set to true to have Amazon Lex treat speech from speakers other than the primary speaker as non-speech. Set to false to disable speaker diarization and rely on voice activity detection alone.
dict
Response Syntax
{
'botId': 'string',
'botVersion': 'string',
'localeId': 'string',
'localeName': 'string',
'description': 'string',
'nluIntentConfidenceThreshold': 123.0,
'voiceSettings': {
'engine': 'standard'|'neural'|'long-form'|'generative',
'voiceId': 'string'
},
'unifiedSpeechSettings': {
'speechFoundationModel': {
'modelArn': 'string',
'voiceId': 'string'
}
},
'audioFillerSettings': {
'enabled': True|False,
'audioType': 'MELODY_CHIPPER_CHIME'|'MELODY_CURIOUS_CRAWL'|'MELODY_RISING_RIPPLE'|'MELODY_PATIENT_PING'|'MELODY_PONDERING_PONG'|'TYPING_KINETIC_KEYS'|'TYPING_QUIET_QWERTY',
'startDelayInMilliseconds': 123,
'minimumPlayDurationInMilliseconds': 123,
'responseDeliveryDelayInMilliseconds': 123
},
'speechRecognitionSettings': {
'speechModelPreference': 'Standard'|'Neural'|'Deepgram'|'Advanced',
'speechModelConfig': {
'deepgramConfig': {
'apiTokenSecretArn': 'string',
'modelId': 'string'
}
}
},
'botLocaleStatus': 'Creating'|'Building'|'Built'|'ReadyExpressTesting'|'Failed'|'Deleting'|'NotBuilt'|'Importing'|'Processing',
'failureReasons': [
'string',
],
'creationDateTime': datetime(2015, 1, 1),
'lastUpdatedDateTime': datetime(2015, 1, 1),
'recommendedActions': [
'string',
],
'generativeAISettings': {
'runtimeSettings': {
'slotResolutionImprovement': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'nluImprovement': {
'enabled': True|False,
'assistedNluMode': 'Primary'|'Fallback'|'DialogManager',
'intentDisambiguationSettings': {
'enabled': True|False,
'maxDisambiguationIntents': 123,
'customDisambiguationMessage': 'string'
}
}
},
'buildtimeSettings': {
'descriptiveBotBuilder': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
},
'sampleUtteranceGeneration': {
'enabled': True|False,
'bedrockModelSpecification': {
'modelArn': 'string',
'guardrail': {
'identifier': 'string',
'version': 'string'
},
'traceStatus': 'ENABLED'|'DISABLED',
'customPrompt': 'string'
}
}
}
},
'speechDetectionSensitivity': 'Default'|'HighNoiseTolerance'|'MaximumNoiseTolerance',
'speakerDiarizationSettings': {
'enabled': True|False
}
}
Response Structure
(dict) --
botId (string) --
The identifier of the bot that contains the updated locale.
botVersion (string) --
The version of the bot that contains the updated locale.
localeId (string) --
The language and locale of the updated bot locale.
localeName (string) --
The updated locale name for the locale.
description (string) --
The updated description of the locale.
nluIntentConfidenceThreshold (float) --
The updated confidence threshold for inserting the AMAZON.FallbackIntent and AMAZON.KendraSearchIntent intents in the list of possible intents for an utterance.
voiceSettings (dict) --
The updated Amazon Polly voice to use for voice interaction with the user.
engine (string) --
Indicates the type of Amazon Polly voice that Amazon Lex should use for voice interaction with the user. For more information, see the engine parameter of the SynthesizeSpeech operation in the Amazon Polly developer guide.
If you do not specify a value, the default is standard.
voiceId (string) --
The identifier of the Amazon Polly voice to use.
unifiedSpeechSettings (dict) --
The updated unified speech settings for the bot locale.
speechFoundationModel (dict) --
The foundation model configuration to use for unified speech processing capabilities.
modelArn (string) --
The Amazon Resource Name (ARN) of the foundation model used for speech processing.
voiceId (string) --
The identifier of the voice to use for speech synthesis with the foundation model.
audioFillerSettings (dict) --
The updated audio filler settings for the bot locale.
enabled (boolean) --
Specifies whether audio filler playback is enabled for the bot locale. Set to true to play filler audio while Amazon Lex processes a user utterance. Set to false to disable filler audio.
audioType (string) --
The identifier of the audio filler to play while Amazon Lex processes the user's input. This field is required when enabled is true.
startDelayInMilliseconds (integer) --
The time, in milliseconds, to wait after the end of the user's utterance before starting audio filler playback. Valid range is 500 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 2500 milliseconds.
minimumPlayDurationInMilliseconds (integer) --
The minimum time, in milliseconds, that audio filler plays once it has started, even if the bot response becomes ready sooner. Valid range is 1000 to 5000 milliseconds. If not specified, Amazon Lex uses a default of 3000 milliseconds.
responseDeliveryDelayInMilliseconds (integer) --
The silent delay, in milliseconds, inserted between the end of audio filler playback and the start of the bot's response. Valid range is 200 to 1000 milliseconds. If not specified, Amazon Lex uses a default of 500 milliseconds.
speechRecognitionSettings (dict) --
The updated speech-to-text settings for the bot locale.
speechModelPreference (string) --
The speech-to-text model to use.
speechModelConfig (dict) --
Configuration settings for the selected speech-to-text model.
deepgramConfig (dict) --
Configuration settings for using Deepgram as the speech-to-text provider.
apiTokenSecretArn (string) --
The Amazon Resource Name (ARN) of the Secrets Manager secret that contains the Deepgram API token.
modelId (string) --
The identifier of the Deepgram speech-to-text model to use for processing speech input.
botLocaleStatus (string) --
The current status of the locale. When the bot status is Built the locale is ready for use.
failureReasons (list) --
If the botLocaleStatus is Failed, the failureReasons field lists the errors that occurred while building the bot.
(string) --
creationDateTime (datetime) --
A timestamp of the date and time that the locale was created.
lastUpdatedDateTime (datetime) --
A timestamp of the date and time that the locale was last updated.
recommendedActions (list) --
Recommended actions to take to resolve an error in the failureReasons field.
(string) --
generativeAISettings (dict) --
Contains settings for generative AI features powered by Amazon Bedrock for your bot locale.
runtimeSettings (dict) --
Contains specifications about the Amazon Lex runtime generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
slotResolutionImprovement (dict) --
An object containing specifications for the assisted slot resolution feature.
enabled (boolean) --
Specifies whether assisted slot resolution is turned on or off.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to assist slot resolution.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
nluImprovement (dict) --
An object containing specifications for the Assisted NLU feature within the bot's runtime settings. These settings determine how the bot processes and interprets user utterances during conversations.
enabled (boolean) --
Determines whether the Assisted NLU feature is enabled for the bot. When set to true, Amazon Lex uses advanced models to improve intent recognition and slot resolution, with the default being false.
assistedNluMode (string) --
Specifies the mode for Assisted NLU operation. Use Primary to make Assisted NLU the primary intent recognition method, Fallback to use it only when standard NLU confidence is low, or DialogManager to use a large language model to manage the conversation, including intent recognition, slot elicitation, and confirmation.
intentDisambiguationSettings (dict) --
An object containing specifications for the Intent Disambiguation feature within the Assisted NLU settings. These settings determine how the bot handles ambiguous user inputs that could match multiple intents.
enabled (boolean) --
Determines whether the Intent Disambiguation feature is enabled. When set to true, Amazon Lex will present disambiguation options to users when multiple intents could match their input, with the default being false.
maxDisambiguationIntents (integer) --
Specifies the maximum number of intent options (2-5) to present to users when disambiguation is needed. This setting determines how many intent options will be shown to users when the system detects ambiguous input. The default value is 3.
customDisambiguationMessage (string) --
Provides a custom message that will be displayed before presenting the disambiguation options to users. This message helps set the context for users and can be customized to match your bot's tone and brand. If not specified, a default message will be used.
buildtimeSettings (dict) --
Contains specifications about the Amazon Lex build time generative AI capabilities from Amazon Bedrock that you can turn on for your bot.
descriptiveBotBuilder (dict) --
An object containing specifications for the descriptive bot building feature.
enabled (boolean) --
Specifies whether the descriptive bot building feature is activated or not.
bedrockModelSpecification (dict) --
An object containing information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
sampleUtteranceGeneration (dict) --
Contains specifications for the sample utterance generation feature.
enabled (boolean) --
Specifies whether to enable sample utterance generation or not.
bedrockModelSpecification (dict) --
Contains information about the Amazon Bedrock model used to interpret the prompt used in descriptive bot building.
modelArn (string) --
The ARN of the foundation model used in descriptive bot building.
guardrail (dict) --
The guardrail configuration in the Bedrock model specification details.
identifier (string) --
The unique guardrail id for the Bedrock guardrail configuration.
version (string) --
The guardrail version for the Bedrock guardrail configuration.
traceStatus (string) --
The Bedrock trace status in the Bedrock model specification details.
customPrompt (string) --
The custom prompt used in the Bedrock model specification details.
speechDetectionSensitivity (string) --
The updated sensitivity level for voice activity detection (VAD) in the bot locale.
speakerDiarizationSettings (dict) --
The updated speaker diarization settings for the bot locale.
enabled (boolean) --
Specifies whether speaker diarization is enabled for the bot locale. Set to true to have Amazon Lex treat speech from speakers other than the primary speaker as non-speech. Set to false to disable speaker diarization and rely on voice activity detection alone.